Federal Patent Jurisdiction and Venue: Section 1338 Arising Under, Section 1400(b) After TC Heartland, PTAB Review, and Federal Circuit Appellate Jurisdiction

Patent litigation is the one area of federal civil practice where jurisdiction, venue, and appellate assignment are all governed by their own statutes, interpreted by their own court, and periodically corrected by the Supreme Court. A patent case cannot be filed in state court, cannot be appealed to the regional circuit, and — since 2017 — cannot be filed against a domestic corporation anywhere except its state of incorporation or a district where it has a regular and established place of business. Each of those propositions has a contested history, and two of them were the law, then were not the law for a generation, then were the law again.

That instability is the practical point of this guide. Patent venue in particular is the rare doctrine where the Supreme Court restored its own precedent after twenty-seven years of contrary Federal Circuit law, redistributing the national patent docket in a single unanimous opinion. Patent appellate jurisdiction is the rare doctrine where Congress overrode the Supreme Court by statute. And the exceptional-case, willfulness, indefiniteness, and laches standards a patent litigator uses today all exist because the Supreme Court replaced a Federal Circuit-specific test with a general one.

This guide covers the grants, the venue analysis as it is actually litigated after TC Heartland, the reviewability of Patent Trial and Appeal Board decisions, the division of appellate labor, and the substantive doctrines that carry jurisdictional or forum consequences.

Contents

• Exclusive subject-matter jurisdiction under § 1338(a)
• Arising under: Christianson, Holmes Group, Gunn
• Venue under § 1400(b): Stonite to TC Heartland
• The post-TC Heartland trilogy: Cray, ZTE, Google
• Foreign defendants, transfer, and joinder
• Declaratory judgment jurisdiction and mootness
• PTAB jurisdiction and reviewability
• Federal Circuit appellate jurisdiction
• The standards the Supreme Court recalibrated
• Substantive doctrines with jurisdictional bite
• Practice checklist
• Frequently asked questions

Exclusive subject-matter jurisdiction under § 1338(a)

28 U.S.C. § 1338(a) gives the district courts original jurisdiction over "any civil action arising under any Act of Congress relating to patents, plant variety protection, copyrights and trademarks," and makes that jurisdiction exclusive of the state courts as to patents, plant variety protection, and copyrights. Trademark is the odd one out: § 1338(a) covers it, but concurrently with the states.

Two consequences follow immediately, and both are litigated.

First, exclusivity is a limit on state courts, not a magnet for every dispute that mentions a patent. A patent license dispute, an assignment dispute, an inventorship-adjacent ownership question resolved by contract, a trade-secret claim about patented technology, a breach of an NDA — all of these routinely belong in state court, because the claim asserted does not arise under the Patent Act. Conversely, a state court that adjudicates infringement or validity as the claim before it has exceeded its power.

Second, because § 1338(a) uses the same "arising under" language as 28 U.S.C. § 1331, the Supreme Court has construed the two in parallel. The well-pleaded complaint rule of Louisville & Nashville Railroad Co. v. Mottley applies: jurisdiction turns on the plaintiff's own properly pleaded claim, not on a defense, not on a counterclaim, and not on the parties' prediction of what will actually decide the case. The embedded-federal-issue framework of Grable & Sons Metal Products, Inc. v. Darue Engineering & Manufacturing also applies, and its leading application is itself a patent case.

The statutory grant of the cause of action is separate. 35 U.S.C. § 281 gives a patentee "remedy by civil action for infringement of his patent," and 35 U.S.C. § 271 defines what infringement is. Section 1338 supplies the forum; §§ 271 and 281 supply the claim. Keep the two separate in pleading — a complaint that cites only § 1338 has pleaded jurisdiction, not a cause of action.

Arising under: Christianson, Holmes Group, Gunn

Three Supreme Court decisions define the boundary, and each did something more than announce a test.

Christianson v. Colt Industries Operating Corp. (1988) is the foundational modern case. The plaintiffs brought antitrust and tortious-interference claims; the theory of the antitrust claim was that Colt's patents were invalid for failure to disclose, so its enforcement conduct was unlawful monopolization. The Federal Circuit and the Seventh Circuit each concluded the other had jurisdiction and transferred the case back and forth. The Supreme Court held that § 1338 jurisdiction exists only where the plaintiff's right to relief "necessarily depends on resolution of a substantial question of federal patent law, in that patent law is a necessary element of one of the well-pleaded claims." Critically, the Court added the rule that still disposes of most of these fights: where a claim is supported by alternative theories, the claim does not arise under patent law "unless patent law is essential to each of those theories." Because Christianson's antitrust claim could succeed on a non-patent theory, the Federal Circuit lacked jurisdiction.

Christianson has a second holding that patent practitioners forget and then need badly. On the ping-pong transfer problem, the Court held that the law-of-the-case doctrine applies to a transferee court's treatment of a transferor court's jurisdictional ruling: a court to which a case is transferred should adhere to the transferring court's decision on jurisdiction unless that decision is "clearly wrong," and "situations might arise in which the transferee court considers the transfer clearly erroneous." The practical instruction was blunt — courts "should be loath to do so" and must not leave a litigant stranded between two courts each disclaiming power. If you draw a jurisdictional transfer between the Federal Circuit and a regional circuit, Christianson is the case that tells the second court to decide the case rather than send it back.

Holmes Group, Inc. v. Vornado Air Circulation Systems, Inc. (2002) applied the well-pleaded complaint rule to appellate jurisdiction with a result Congress did not accept. Holmes filed a declaratory action about trade dress; Vornado counterclaimed for patent infringement. The Federal Circuit took the appeal. The Supreme Court reversed, holding that a patent-law counterclaim does not make the case one "arising under" patent law for purposes of the Federal Circuit's jurisdiction, because the plaintiff is master of the complaint and a counterclaim cannot create arising-under jurisdiction any more than a defense can. The immediate consequence was forum-shopping in reverse: a declaratory plaintiff could keep a patent dispute out of the Federal Circuit by structuring the pleadings.

That holding is no longer the law, and the reason is worth knowing. In the Leahy-Smith America Invents Act of 2011 Congress abrogated Holmes Group directly. It rewrote 28 U.S.C. § 1295 to give the Federal Circuit jurisdiction over appeals in "any civil action arising under, or in any civil action in which a party has asserted a compulsory counterclaim arising under, any Act of Congress relating to patents," and it added 28 U.S.C. § 1454, a patent-specific removal provision permitting removal of a "civil action in which any party asserts a claim for relief arising under any Act of Congress relating to patents" — removable by "any party," on a clock the district court may extend for cause, and with the ordinary § 1447(d) bar on reviewing remands expressly displaced for § 1454 removals.

This is a clean, teachable example of Congress correcting the Supreme Court rather than the reverse. The Court read the statute as written; Congress disagreed with the policy result and amended the statute. When you brief a Federal Circuit jurisdictional question today, cite Holmes Group for the well-pleaded complaint principle, then cite the amended § 1295(a)(1) and § 1454 for the rule that actually governs counterclaims. Citing Holmes Group for its result is an error.

Gunn v. Minton (2013) closed the other flank. Minton lost a patent infringement suit because his attorneys failed to raise the experimental-use exception to the on-sale bar, and sued them for malpractice in Texas state court. He then argued — to escape an adverse state judgment — that the malpractice claim arose under patent law and belonged exclusively in federal court under § 1338(a). A unanimous Court held it did not, and used the occasion to restate Grable as four conjunctive requirements. The federal issue must be (1) necessarily raised, (2) actually disputed, (3) substantial, and (4) capable of resolution in federal court without disrupting the federal-state balance.

The malpractice claim satisfied the first two: the state court would have to decide the hypothetical patent question of whether the experimental-use argument would have succeeded. It failed the third and fourth. "Substantial" means substantial to the federal system as a whole, not merely important to the litigants, and the patent question here was "backward-looking" and hypothetical — no actual patent would be granted, invalidated, or enforced differently by the state court's answer, and the resolution would not bind anyone in a future federal patent case. The Court also emphasized the federalism side: states have a "special responsibility for maintaining standards among members of the licensed professions," and nothing in the federal interest in uniform patent law is served by pulling legal-malpractice litigation into federal court.

Gunn is now the controlling framework for every "patent-adjacent" state-law claim. The recurring winners for state jurisdiction are malpractice, license and royalty disputes, assignment and ownership disputes governed by contract, and employment claims about invention-assignment agreements. The recurring loser is a claim drafted so that infringement or validity is the claim itself rather than an ingredient.

Venue under § 1400(b): Stonite to TC Heartland

28 U.S.C. § 1400(b) is eight lines long and has generated more forum litigation than any other venue provision in the Code:

Any civil action for patent infringement may be brought in the judicial district where the defendant resides, or where the defendant has committed acts of infringement and has a regular and established place of business.

Two prongs, disjunctive. Residence, or acts of infringement plus a regular and established place of business. The entire history of patent venue is a fight over what "resides" means and whether the general venue statute, 28 U.S.C. § 1391, supplements § 1400(b).

Stonite (1942): § 1400(b) is exclusive

Stonite Products Co. v. Melvin Lloyd Co. held that the patent venue provision — then § 48 of the Judicial Code — is the exclusive provision governing venue in patent infringement actions, and is not supplemented by the general venue statute. The plaintiff had sued two Pennsylvania defendants in the Western District of Pennsylvania, where only one of them resided, relying on the general multi-defendant venue provision. The Court refused, holding that the patent venue statute "alone should control venue in patent infringement proceedings" and that Congress had enacted it as "a special venue provision" precisely to define the whole of patent venue. That holding — exclusivity — is the load-bearing premise for everything that followed.

Fourco (1957): "resides" means state of incorporation

The 1948 recodification split the old provision into § 1400(b) and added a definitional subsection to the general statute, § 1391(c), providing that a corporation "may be sued in any judicial district in which it is incorporated or licensed to do business or is doing business, and such judicial district shall be regarded as the residence of such corporation for venue purposes."

Fourco Glass Co. v. Transmirra Products Corp. held that this changed nothing for patent cases. "Resides" in § 1400(b), as applied to a domestic corporation, means the state of incorporation and nothing else. The Court gave two reasons that recur in the later cases. First, § 1400(b) "is complete, independent and alone controlling in its sphere" — the Stonite exclusivity holding survived the recodification. Second, the 1948 revision was not understood to make substantive changes, and "no changes of law or policy are to be presumed from changes of language in the revision unless an intent to make such changes is clearly expressed." The Court also invoked the specific-over-general canon: § 1391(c) is a general provision, § 1400(b) a specific one, and "the specific statute will not be controlled or nullified by a general one."

VE Holding (1990): the Federal Circuit reads it the other way

In 1988 Congress amended § 1391(c) again, this time to read: "For purposes of venue under this chapter, a defendant that is a corporation shall be deemed to reside in any judicial district in which it is subject to personal jurisdiction at the time the action is commenced."

VE Holding Corp. v. Johnson Gas Appliance Co. held that the new opening phrase — "For purposes of venue under this chapter" — was the clearly expressed intent Fourco said was missing. Section 1400(b) sits in the same chapter as § 1391. Therefore, the Federal Circuit reasoned, the 1988 definition of corporate residence applies to § 1400(b), and a corporate defendant "resides" in any district where it is subject to personal jurisdiction. The court acknowledged it was displacing Fourco and justified doing so on the ground that Congress had changed the statutory text Fourco construed.

The practical effect was to collapse patent venue into personal jurisdiction. Any accused infringer selling nationwide could be sued in any of the ninety-four districts. Plaintiffs chose accordingly, and the consequences were extreme and well documented: by the mid-2010s the Eastern District of Texas alone was receiving roughly a quarter to nearly half of all patent cases filed in the United States in a given year, with a single judge's docket at times exceeding that of most circuits' entire patent caseload. Local practice — fast schedules, disfavor toward early summary judgment and transfer, high plaintiff win rates before juries — became a substantive input into patent value. VE Holding stood for twenty-seven years.

TC Heartland (2017): Fourco restored

TC Heartland LLC v. Kraft Foods Group Brands LLC reversed, unanimously, in eight and a half pages. TC Heartland was organized under Indiana law and headquartered there; Kraft sued it in Delaware. The Federal Circuit denied mandamus on the authority of VE Holding.

The Supreme Court held that Fourco remains the law. The reasoning is entirely about the burden of showing a statutory change:

• Fourco held "resides" in § 1400(b) means state of incorporation for a domestic corporation. That is a definitive interpretation of the patent venue statute, which Congress has not amended.
• The 1988 and 2011 changes were to § 1391(c), not to § 1400(b). The current § 1391(a) makes the general statute applicable "[e]xcept as otherwise provided by law," which if anything strengthens the case that a specific venue statute governs on its own terms.
• The "for purposes of venue under this chapter" language does not carry the weight VE Holding placed on it, because nothing in it suggests Congress meant to ratify a reading of § 1391(c) that had already been rejected as to § 1400(b). Congress's 2011 addition of the saving clause "cannot be viewed as a ratification" of the Federal Circuit's contrary construction.

The holding: § 1400(b) "is not supplemented by § 1391(c)," and a domestic corporation "resides" only in its state of incorporation. Justice Thomas wrote for eight Justices; Justice Gorsuch did not participate as to one issue and Justice Thomas noted the Court expressed no view on the venue rules for unincorporated entities or foreign defendants.

This is the structural point worth pausing on. The Supreme Court did not overrule itself in TC Heartland; it restored its own precedent that a lower court had displaced. The relevant lesson for briefing is that a court of appeals reading a Supreme Court statutory holding as abrogated by an amendment to a different statute is on thin ice, and the "clearly expressed intent" standard from Fourco is a real one.

The immediate practical consequence

The redistribution was fast and large. Filings in the Eastern District of Texas dropped precipitously in the months after May 2017; Delaware — the state of incorporation for an enormous share of American operating companies — became the single largest patent venue, and the Northern District of California and District of New Jersey absorbed much of the remainder. Within a few years the Western District of Texas emerged as a new concentration, but on a different theory: not residence, but the second prong, because most large technology companies maintain offices in Austin.

That is the durable change. Before TC Heartland, the first prong swallowed everything and the "regular and established place of business" prong was nearly dead law, unexamined since the 1980s. After TC Heartland it became the entire battleground, and the Federal Circuit had to build the doctrine essentially from scratch.

The post-TC Heartland trilogy: Cray, ZTE, Google

Three Federal Circuit mandamus decisions built the second prong. Together they are the framework every § 1400(b) motion is now argued under.

In re Cray: the three requirements

In re Cray Inc. (Fed. Cir. 2017) arose four months after TC Heartland. Cray, a Washington corporation, was sued in the Eastern District of Texas. It had no offices or property in the district. It had two sales employees who worked from home there, one of whom had generated roughly $345 million in sales over seven years and was reimbursed for internet and phone. The district court denied transfer under a four-factor test of its own devising keyed to "physical presence, defendant's representations, benefits received, and targeted interactions."

The Federal Circuit granted mandamus and replaced that test with a textual one. Section 1400(b)'s second prong requires three things, each necessary:

1. A physical place in the district. Not merely a virtual space or the ability to access an electronic database. "The statute cannot be read to refer merely to a virtual space or to electronic communications from one person to another." A "place" means "a building or a part of a building set apart for any purpose or quarters of any kind from which business is conducted."
2. A regular and established place of business. "Regular" means it operates in a "steady, uniform, orderly, and methodical manner" — sporadic activity does not suffice. "Established" means "settled certainly, or fixed permanently"; a place of business that is transient or that the defendant can and does move at will is not established. Five years of continuous presence is the paradigm; a single trade-show booth is not.
3. The place of the defendant. This is where most motions are decided. The place must be "the place of the defendant, not solely a place of the defendant's employee." Relevant considerations: whether the defendant owns, leases, or exercises "other attributes of possession or control over the place"; whether the defendant conditioned employment on maintaining a location in the district; whether the defendant lists the address in advertising or on its website; and whether it stores inventory or materials there for distribution or sale.

Applying that, the home offices of two telecommuting employees were not Cray's places of business. Cray did not own, lease, rent, or pay for the homes; it did not require the employees to live in the district; it never advertised the homes as its own locations. That the employees' work was valuable and that Cray reimbursed some expenses was irrelevant to possession or control. The court added the point that decides most remote-work disputes: employees "can move" their home offices, and a place a defendant has no ratified interest in is not the defendant's.

Cray is the first citation in virtually every patent venue brief filed since. Note also its holding on the scope of the inquiry: the court rejected any test that "is not sufficiently tethered to [the] statutory language," including tests borrowed from personal-jurisdiction minimum-contacts analysis. Venue is not contacts.

In re ZTE: Federal Circuit law and the burden

In re ZTE (USA) Inc. (Fed. Cir. 2018) answered two questions Cray had left open, and they are procedural questions that decide cases.

First, which body of law governs. The district court had applied Fifth Circuit venue law, under which the plaintiff's venue allegations are taken as true if not specifically contradicted. The Federal Circuit held that the § 1400(b) analysis is "unique to patent law" and therefore governed by Federal Circuit law, not regional circuit law. That matters because it forecloses the argument — still occasionally made — that a favorable regional-circuit venue presumption applies.

Second, who bears the burden. The Federal Circuit held that upon a motion to dismiss for improper venue, the plaintiff bears the burden of establishing proper venue. The court reasoned from the general rule that a plaintiff who chooses a forum must justify it and from § 1400(b)'s status as a restrictive statute entitled to narrow construction. On the merits, ZTE's relationship with a third-party call center in the district — which handled customer service calls under contract — was insufficient on the record, and the court remanded for application of the correct burden and the Cray factors.

Practically: a defendant's venue motion should be short. Deny the elements, put in a declaration establishing no lease, no ownership, no employment condition, no advertised address, and let the plaintiff carry the load. A plaintiff facing ZTE needs affirmative evidence of possession or control, and needs it early, because venue discovery is discretionary and often narrow.

In re Google: employees or agents, not servers

In re Google LLC (Fed. Cir. 2020) resolved the highest-stakes open question of the post-TC Heartland era. Google had no offices in the Eastern District of Texas. It did have Google Global Cache servers physically located in the facilities of local internet service providers, under agreements by which the ISPs housed and maintained the hardware and Google retained ownership and exclusive control of the equipment and rack space. The plaintiffs argued the leased rack space was a physical place, regular and established, and Google's.

The Federal Circuit granted mandamus and held venue improper, on a ground that narrowed the prong considerably: a "regular and established place of business" requires the regular, physical presence of an employee or other agent of the defendant conducting the defendant's business at the alleged place. The court reached that reading from the statutory history — the "place of business" requirement descends from provisions understood to require a physical agent — and from the phrase "place of business" itself, which "connotes a place where business is carried on," not merely where equipment sits. The ISP personnel who installed and maintained the servers were not Google's agents conducting Google's business; their contractual duties were limited to hosting and basic maintenance, which is not "conducting" Google's business of serving content.

Two caveats worth carrying into a brief. The panel expressly declined to decide whether a "place of business" also requires that the business be conducted with customers physically present, and it noted the mandamus posture — relief was appropriate because the question was "basic and undecided" and recurring. Later district courts have generally treated the employee-or-agent requirement as holding, not dictum.

The combined effect of the trilogy is a genuinely restrictive second prong. Server farms, rack space, cloud infrastructure, distribution through unaffiliated retailers, third-party call centers, and remote employees' homes ordinarily do not create venue. Leased or owned offices, staffed retail locations, warehouses with company personnel, and addresses the defendant advertises as its own ordinarily do.

Foreign defendants, transfer, and joinder

Foreign defendants

TC Heartland expressly reserved the question of venue over foreign defendants, and the answer comes from § 1391 rather than from § 1400(b)'s residence prong. Section 1391(c)(3) provides that a defendant not resident in the United States "may be sued in any judicial district, and the joinder of such a defendant shall be disregarded in determining where the action may be brought with respect to other defendants." The Federal Circuit has long applied that principle to patent cases, and it survives TC Heartland: a foreign corporation not resident in any district may be sued in any district.

The reasoning is not that § 1391 supplements § 1400(b) — TC Heartland forbids that. It is that a foreign entity has no state of incorporation within the United States, so the residence prong of § 1400(b) has no domestic referent, and § 1391(c)(3) supplies the default. The practical consequence is asymmetry that shapes case structure: a plaintiff suing a foreign manufacturer and its U.S. subsidiary must establish venue over the subsidiary under § 1400(b), but faces no venue obstacle as to the parent. Defendants respond by contesting whether the entity is genuinely foreign — a U.S.-incorporated subsidiary of a foreign parent is a domestic corporation, and TC Heartland applies to it in full.

Two related points. First, alter-ego and agency theories are litigated to impute a subsidiary's place of business to a foreign or out-of-state parent; the Federal Circuit requires the corporate-separateness showing that would support veil-piercing, not merely common ownership or a shared brand. Second, venue for unincorporated entities — LLCs, partnerships, LLPs — was also reserved in TC Heartland, and the district courts have largely settled on treating an LLC as residing where it is organized, by analogy to the Fourco rule, rather than borrowing the citizenship rule that traces every member.

Transfer under § 1404(a)

The most important practical development after TC Heartland is that the venue fight moved from § 1400(b) to § 1404(a). Once the concentration shifted to districts where large defendants genuinely have offices — Austin and the Western District of Texas being the paradigm — the argument is no longer "venue is improper" but "this district is inconvenient."

Section 1404(a) permits transfer "[f]or the convenience of parties and witnesses, in the interest of justice" to any other district where the action might have been brought or to which all parties consent. In patent cases the transferee court's venue propriety is judged under § 1400(b), so the Cray analysis still operates as a constraint on where a case can go. The convenience analysis itself is governed by regional circuit law, not Federal Circuit law — the opposite of the § 1400(b) rule from ZTE — because § 1404(a) is a general procedural statute.

What is distinctively patent about § 1404(a) practice is the Federal Circuit's willingness to issue mandamus on transfer denials. Beginning with the Volkswagen line in the Fifth Circuit and the Federal Circuit's TS Tech and Genentech decisions, and intensifying in the Western District of Texas period, the Federal Circuit has repeatedly granted mandamus where a district court gave inadequate weight to the location of documents and evidence, the relative ease of access to sources of proof, the availability of compulsory process over non-party witnesses, or the local interest, or where it overweighted its own docket speed and the plaintiff's choice of forum. That is unusual — mandamus on a discretionary case-management ruling is ordinarily unavailable — and it is the single most effective tool a patent defendant has once venue is proper.

Preserve the record accordingly. A § 1404(a) motion should identify named non-party witnesses and their locations, the physical location of the accused-product engineering documents and the engineers who created them, and the absence of any connection between the chosen district and the accused design. A plaintiff opposing should build the local-interest and party-witness record early, because the Federal Circuit reviews the record the district court had.

Joinder limits under § 299

The AIA addressed forum concentration from a second direction. 35 U.S.C. § 299 provides that accused infringers may be joined in one action, or have their actions consolidated for trial, only if the allegations arise out of "the same transaction, occurrence, or series of transactions or occurrences relating to the making, using, importing into the United States, offering for sale, or selling of the same accused product or process" — and it expressly forecloses the practice it was enacted to stop: "accused infringers may not be joined in one action as defendants or counterclaim defendants, or have their actions consolidated for trial, based solely on allegations that they each have infringed the patent or patents in suit."

Before § 299, a plaintiff could name twenty unrelated defendants in a single complaint in a favored district, generating settlement pressure at minimal filing cost. Section 299 ended that. Note two things about its operation. First, it is waivable — a defendant may consent to joinder, and § 299(c) says so expressly. Second, it governs joinder and consolidation for trial, not pretrial consolidation, so district courts routinely coordinate discovery and claim construction across separately filed cases against unrelated defendants while keeping trials separate. The practical consequence of § 299 plus TC Heartland is that a campaign against multiple defendants now means multiple complaints in multiple districts, which is precisely the cost Congress and the Court intended to reimpose.

Declaratory judgment jurisdiction and mootness

Accused infringers and licensees often want to be plaintiffs. 28 U.S.C. § 2201 permits a declaratory judgment "[i]n a case of actual controversy within its jurisdiction," which imports Article III's case-or-controversy requirement and nothing more. The Declaratory Judgment Act creates a remedy, not jurisdiction; jurisdiction over a declaratory action seeking a judgment of invalidity or noninfringement comes from § 1338(a), because the coercive action it displaces would be an infringement suit.

MedImmune, Inc. v. Genentech, Inc. (2007) is the governing case and it displaced a Federal Circuit-specific rule. MedImmune was paying royalties under a license and wanted to challenge the licensed patent's validity without breaching. The Federal Circuit dismissed for lack of a case or controversy under its "reasonable apprehension of imminent suit" test — a licensee in good standing faces no apprehension of suit, because the license is a defense.

The Supreme Court reversed. The question is whether "the facts alleged, under all the circumstances, show that there is a substantial controversy, between parties having adverse legal interests, of sufficient immediacy and reality to warrant the issuance of a declaratory judgment." A licensee need not breach or repudiate the license to sue. The Court drew on the coerced-payment line — a party forced to choose between abandoning its rights and risking treble damages and an injunction has a justiciable controversy — and rejected the argument that the license was a covenant not to challenge validity absent express language. Footnote 11 did the doctrinal work most directly, observing that the reasonable-apprehension test conflicted with the Court's precedents and could not stand as the exclusive standard.

The consequences for practice were immediate and remain: a demand letter, a licensing overture accompanied by claim charts, a course of conduct putting a competitor's product in issue, or an existing license under protest can each support a declaratory action. The Federal Circuit's post-MedImmune framework asks whether the patentee's conduct created a definite and concrete dispute about a specific accused product, which is a real limit — a general assertion that a portfolio "may be relevant" to an industry usually is not enough — but it is far looser than the test it replaced. Note also that § 2201's "may declare" is discretionary: a district court may decline a declaratory action, and duplicative first-filed declaratory suits are a recurring target.

Already, LLC v. Nike, Inc. (2013) supplies the other half of the analysis — how a patentee can end the controversy. Nike sued Already for trademark infringement, Already counterclaimed to invalidate the mark, and Nike then issued a broad covenant not to sue and dismissed. The Court held the counterclaim moot. It is a trademark case, but the mootness framework governs IP declaratory practice generally and is cited constantly in patent cases.

The framework: the voluntary cessation doctrine places a "formidable burden" on the party asserting mootness to show "it is absolutely clear the allegedly wrongful behavior could not reasonably be expected to recur." A covenant not to sue can carry that burden if it is unconditional and irrevocable, covers past and future conduct, reaches the defendant's distributors and customers, and extends to any colorable imitation of the accused product. Nike's did. The Court rejected Already's argument that its interest in competing generally or investors' concerns kept the case alive, holding that speculative future products and generalized market injury do not preserve a controversy.

The drafting lesson is precise. A patentee seeking to moot an invalidity counterclaim needs a covenant that is unconditional, irrevocable, forward- and backward-looking, and product-scope-generous. A covenant limited to "current versions" of the accused product, or reserving rights against downstream customers, will not do it — and a failed attempt leaves the patentee litigating validity with its own covenant in evidence.

PTAB jurisdiction and reviewability

The AIA created a parallel forum for validity challenges, and the resulting question — how much of what the Patent Trial and Appeal Board does is reviewable in court — has produced five Supreme Court decisions in six years.

The statutory architecture is short and consequential:

The pattern across the cases is that the institution side is largely insulated from review while the final written decision side is reviewable — and that the Court has read "institution" broadly.

Cuozzo Speed Technologies, LLC v. Lee (2016) decided both of the foundational questions. On reviewability, the Court held that § 314(d)'s "final and nonappealable" language bars judicial review of the institution decision, at least where the challenge "consists of questions that are closely tied to the application and interpretation of statutes related to the Patent Office's decision to initiate inter partes review." Cuozzo's complaint — that the petition failed § 312(a)(3)'s particularity requirement as to one claim — was exactly that, and unreviewable even after final decision. The Court reserved review for constitutional questions and for actions that exceed statutory limits in ways unrelated to the institution standard. On the second question, the Court upheld the PTO's regulation applying the broadest reasonable interpretation claim-construction standard in IPR under Chevron, reasoning that § 316(a)(4) delegated rulemaking authority and that IPR is not simply a substitute for district court litigation. (The PTO has since changed that standard by regulation to the Phillips standard used in district court; Cuozzo's holding is that the agency had the authority to choose, which is why the change was lawful.)

SAS Institute Inc. v. Iancu (2018) cut the other way and is the reason IPR practice looks as it does. The PTO's regulation permitted partial institution — the Board could institute on some challenged claims and decline as to others, issuing a final written decision addressing only the instituted claims. The Court held this unlawful. Section 318(a) requires the Board to "issue a final written decision with respect to the patentability of any patent claim challenged by the petitioner," and "any" means every one. The petitioner, not the Director, defines the scope of the proceeding: "Congress chose to structure a process in which it's the petitioner, not the Director, who gets to define the contours of the proceeding." Section 314(d)'s nonreviewability did not save the practice, because the challenge was not to the decision whether to institute but to the Board's obligation once it did.

The practical result was all-or-nothing institution, which in turn increased the importance of discretionary denial and of the Director's institution policy. It also closed a mismatch that had bedeviled estoppel: with partial institution, claims challenged but not instituted fell into an estoppel gap.

Thryv, Inc. v. Click-To-Call Technologies, LP (2020) extended Cuozzo to the § 315(b) time bar, which many practitioners had assumed was reviewable because it is a statutory condition external to the merits. The Court held otherwise: a § 315(b) ruling is part of "the Patent Office's decision to initiate inter partes review," and § 314(d) therefore bars appeal. The Court emphasized the AIA's purpose of weeding out bad patents efficiently — allowing appeals on the time bar would "operate to save bad patent claims" by unwinding a completed merits adjudication on a threshold ground. The dissent's structural objection, that § 315(b) protects a patent owner's interest in repose that the majority leaves unenforceable, is the argument to make if you are trying to distinguish Thryv, and the narrow openings are constitutional challenges and claims that the Board acted "outside its statutory limits" in a way untethered to institution.

Oil States Energy Services, LLC v. Greene's Energy Group, LLC (2018) answered the existential challenge. Oil States argued that because a patent is a property right, only an Article III court with a jury may cancel it. The Court disagreed, 7-2. The grant of a patent is "a matter involving public rights" — specifically, "the grant of a public franchise" — and reconsideration of that grant by the agency that made it falls within the public-rights exception to Article III adjudication. The Seventh Amendment argument failed for the same reason: where Congress may properly assign a matter to an administrative tribunal, no jury right attaches. The Court was careful to cabin the holding: it addressed only the "precise constitutional challenges" presented, expressly reserving retroactivity (patents issued before the AIA), due process challenges, and takings claims. Those reservations have generated litigation but not a successful challenge.

Return Mail, Inc. v. United States Postal Service (2019) is a statutory-standing decision with practical reach. The AIA review provisions permit a "person" to petition. The Court held 6-3 that the federal government is not a "person" eligible to petition for IPR, post-grant review, or covered business method review. The longstanding interpretive presumption that "person" excludes the sovereign applied, and the Postal Service could not overcome it — the Court noted that the government's exposure is limited to compensation under 28 U.S.C. § 1498 rather than injunctions and damages, so it is not similarly situated to a private accused infringer. The consequence: federal agencies accused of using patented technology must defend in the Court of Federal Claims under § 1498 and cannot use the Board to invalidate, though they may of course raise invalidity as a defense there.

United States v. Arthrex, Inc. (2021) completes the set and must be discussed without a link here, because the opinion did not resolve in our corpus. Arthrex argued that the Board's administrative patent judges are principal officers who must be appointed by the President with Senate confirmation, not by the Secretary of Commerce. A fractured Court agreed that the appointments were unconstitutional as applied, on the ground that APJs exercised unreviewable authority to issue final decisions on behalf of the Executive Branch while being insulated from Director control — "the unreviewable authority wielded by APJs during inter partes review is incompatible with their appointment by the Secretary to an inferior office."

The remedy is the part that governs practice. Rather than invalidating the IPR regime or requiring Senate confirmation of every APJ, the Court severed the statutory limits on the Director's review authority and held that the Director may review final written decisions on his or her own initiative. That produced the Director Review procedure the PTO implemented immediately afterward: a party dissatisfied with a final written decision may request Director review, and the Director may also act sua sponte. Two operational consequences follow. First, Director review is now an interposed step between the Board and the Federal Circuit that a party should ordinarily consider, both because it can fix the decision and because it shapes the record. Second, the Appointments Clause objection is preserved only for parties whose decisions predated the remedy and who raised it; the Federal Circuit's remand practice under Arthrex required a properly preserved challenge.

Federal Circuit appellate jurisdiction

28 U.S.C. § 1295(a)(1) gives the Federal Circuit exclusive jurisdiction over an appeal from a final decision of a district court "in any civil action arising under, or in any civil action in which a party has asserted a compulsory counterclaim arising under, any Act of Congress relating to patents or plant variety protection." Section 1295 also carries the appeal from the PTAB under § 319, from the ITC, and from the Court of Federal Claims.

Three features matter in practice.

Mixed cases go entirely to the Federal Circuit. If a case arises under the patent laws, the Federal Circuit takes the whole appeal — patent claims, state-law claims, antitrust claims, everything — and applies regional circuit law to the non-patent issues while applying its own law to patent-specific ones. There is no splitting of a single judgment between two courts of appeals. That is why the Christianson alternative-theories rule is jurisdictionally consequential: it determines whether an antitrust appeal with a patent ingredient goes to the Federal Circuit or to the regional circuit.

A counterclaim now suffices. As discussed above, the AIA's amendment to § 1295(a)(1) abrogated Holmes Group. A compulsory patent counterclaim routes the appeal to the Federal Circuit even where the complaint pleads only state-law or non-patent federal claims. Note the qualifier — "compulsory" — which imports Rule 13(a): a permissive patent counterclaim arguably does not trigger § 1295, a question that arises rarely but is worth flagging when the counterclaim is unrelated to the complaint's transaction.

Section 1454 removal is the front-end companion. 28 U.S.C. § 1454 allows removal of any civil action in which any party asserts a claim for relief arising under the patent laws. Three departures from ordinary removal practice: the removing party need not be a defendant; the rule of unanimity does not apply; and the district court may extend the removal deadline "for cause." Section 1454(d)(2) also displaces the § 1447(d) bar, so a remand of a § 1454-removed case is reviewable. The combination means a state-court defendant facing a patent counterclaim dispute has a federal route it lacked before 2011.

Claim construction and the standard of review

Two Supreme Court decisions define how the Federal Circuit reviews what it reviews most.

Markman v. Westview Instruments, Inc. (1996) held unanimously that the construction of patent claims, "including terms of art within its claim, is exclusively within the province of the court" and not the jury. The Seventh Amendment does not require otherwise, because claim construction is a "mongrel practice" historically performed by judges and because the need for uniformity in patent construction — the "importance of uniformity in the treatment of a given patent" — favors judicial resolution. That holding created the Markman hearing, the single most consequential pretrial event in most patent cases, and it made claim construction the principal vehicle for Federal Circuit control of patent outcomes.

Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc. (2015) narrowed that control. The Federal Circuit had reviewed claim construction entirely de novo, treating it as a pure question of law under Markman, including the subsidiary findings about how a skilled artisan would understand a term. The Supreme Court held that Rule 52(a)(6) admits no exception for patent claim construction: "when reviewing a district court's resolution of subsidiary factual matters made in the course of its construction of a patent claim, the Federal Circuit must apply a 'clear error,' not a 'de novo,' standard of review."

The line Teva draws is workable and heavily litigated. Construction based solely on the intrinsic evidence — the claims, specification, and prosecution history — is a legal determination reviewed de novo. Findings resting on extrinsic evidence — expert testimony about the meaning of a term of art in the field, the state of the art, industry usage — are factual and reviewed for clear error, though the ultimate construction remains legal. The practical instruction for trial counsel is to build an extrinsic-evidence record and secure explicit factual findings if you win below, and to frame the dispute as purely intrinsic if you intend to appeal.

Dennison Manufacturing Co. v. Panduit Corp. (1986) is the earlier instance of the same correction and belongs in any Teva brief. The Federal Circuit had reversed a district court's nonobviousness holding without addressing the district court's findings under Rule 52(a). The Supreme Court vacated in a short per curiam opinion, directing the Federal Circuit to explain whether Rule 52(a) had been properly applied to the subsidiary findings on the Graham factors. Dennison establishes that the Rule 52(a) point is not a 2015 innovation: the Court had told the Federal Circuit the same thing twenty-nine years earlier, and Teva is best read as an enforcement of Dennison rather than a departure from settled practice.

The standards the Supreme Court recalibrated

Between 2014 and 2017 the Supreme Court dismantled a series of Federal Circuit-specific tests and replaced each with a general standard drawn from ordinary civil practice. Read individually these are discrete doctrinal fixes. Read together they are a sustained statement about the relationship between a specialized court and the general law, and they are the reason a patent litigator's toolkit looks the way it does.

Octane Fitness, LLC v. Icon Health & Fitness, Inc. (2014) addressed fee-shifting. 35 U.S.C. § 285 permits a court to "award reasonable attorney fees to the prevailing party" in "exceptional cases." The Federal Circuit had built a rigid two-part test requiring either litigation-related misconduct of an independently sanctionable magnitude, or a showing that the litigation was both brought in subjective bad faith and objectively baseless, proved by clear and convincing evidence. The Court rejected the whole structure unanimously. "Exceptional" carries its ordinary meaning: "simply one that stands out from others with respect to the substantive strength of a party's litigating position (considering both the governing law and the facts of the case) or the unreasonable manner in which the case was litigated." District courts decide case-by-case "considering the totality of the circumstances." And the Federal Circuit's clear-and-convincing standard was wrong — § 285 "demands a simple discretionary inquiry" governed by the ordinary preponderance standard.

Highmark Inc. v. Allcare Health Management System, Inc. (2014), decided the same day, supplied the review standard. Because Octane Fitness makes the § 285 determination a discretionary one committed to the district court, the Federal Circuit must review it for abuse of discretion, not de novo — including the "objective baselessness" component it had previously reviewed without deference. The pair is the model for the whole line: replace a specialized rigid test with a discretionary standard, then insulate the exercise of that discretion from de novo appellate substitution.

Halo Electronics, Inc. v. Pulse Electronics, Inc. (2016) did the same to enhanced damages. 35 U.S.C. § 284 permits a court to "increase the damages up to three times the amount found or assessed." The Federal Circuit's Seagate framework required a patentee to prove by clear and convincing evidence that the infringer acted despite an objectively high likelihood of infringement, and only if that objective prong was satisfied could the subjective prong be reached — with the objective prong reviewed de novo. The Court struck it down unanimously, reasoning in explicit parallel to Octane Fitness. The objective threshold "excludes from discretionary punishment many of the most culpable offenders," including the deliberate pirate who happens to find a plausible defense later in litigation. Culpability is measured by "the actor's knowledge at the time of the challenged conduct." The evidentiary standard is preponderance, and review is for abuse of discretion. The Court preserved the essential limit: enhanced damages remain reserved for "egregious cases of misconduct beyond typical infringement" and are not a routine consequence of losing.

Nautilus, Inc. v. Biosig Instruments, Inc. (2014) replaced the indefiniteness standard. 35 U.S.C. § 112(b) requires claims that "particularly point[] out and distinctly claim[]" the invention. The Federal Circuit had held a claim indefinite only if it was "insolubly ambiguous" or "not amenable to construction." The Court held that standard "does not satisfy the statute's definiteness requirement" and adopted a new one: a patent is invalid for indefiniteness if its claims, "read in light of the specification delineating the patent, and the prosecution history, fail to inform, with reasonable certainty, those skilled in the art about the scope of the invention." The Court's stated concern was the notice function — the "insolubly ambiguous" formulation "can leave courts and the patent bar at sea without a reliable compass" and tolerated "the innovation-discouraging zone of uncertainty" against which the definiteness requirement guards.

SCA Hygiene Products Aktiebolag v. First Quality Baby Products, LLC (2017) eliminated laches as a defense to damages within the statutory period. 35 U.S.C. § 286 bars recovery for infringement "committed more than six years prior to the filing of the complaint." The Federal Circuit, sitting en banc, had preserved laches as an additional equitable bar within that window. The Court reversed 7-1, applying its reasoning from Petrella v. Metro-Goldwyn-Mayer, Inc.: laches "cannot be invoked to bar legal relief" where Congress has enacted a statute of limitations, because doing so lets courts override a legislative judgment about timeliness. The separation-of-powers framing is explicit — applying laches within § 286 would "give judges a 'legislation-overriding' role." Equitable estoppel survives and is now the operative delay defense; unlike laches it requires misleading conduct, reliance, and prejudice, and when established it bars the claim entirely rather than merely trimming damages.

The pattern, stated plainly: the Supreme Court repeatedly found that the Federal Circuit had converted a discretionary or standard-like statutory command into a rigid multi-part test with an elevated burden of proof and de novo review, and in each case restored the ordinary approach — text-driven meaning, totality-of-circumstances discretion, preponderance, and deference to the trial court. When you encounter a Federal Circuit test with a bright-line threshold, an elevated evidentiary standard, and de novo review, treat its durability as an open question.

Substantive doctrines with jurisdictional bite

Several merits doctrines determine who can be sued, where, and for what — which makes them part of the forum analysis even though they are nominally substantive. 35 U.S.C. § 271 defines infringement and 35 U.S.C. § 281 supplies the civil action; the following decisions define their reach.

Limelight Networks, Inc. v. Akamai Technologies, Inc. (2014) held that there can be no inducement liability under § 271(b) without a single act of direct infringement under § 271(a). The Federal Circuit, en banc, had held Limelight liable for inducement where it performed most steps of a method claim and instructed its customers to perform the remaining step — no single actor performed every step, so there was no direct infringement, but the court found inducement anyway. The Supreme Court reversed unanimously: "liability for inducement must be predicated on direct infringement," and a method patent is not infringed "unless all the steps are carried out." The Court declined to revisit the Federal Circuit's single-entity rule from Muniauction, and on remand the Federal Circuit broadened attribution — conduct of another is attributable where the accused directs or controls that performance or where the actors form a joint enterprise. The forum consequence is real: divided-infringement cases require a defendant against whom every step can be attributed, which constrains whom a plaintiff can name and therefore where it can establish venue.

WesternGeco LLC v. ION Geophysical Corp. (2018) addressed extraterritorial damages. Section 271(f)(2) makes it infringement to supply from the United States a component specially adapted for use in a patented invention, intending it be combined abroad. ION supplied components abroad where they were assembled into infringing survey systems, and WesternGeco proved it lost ten specific survey contracts — roughly $93 million in profits earned outside the United States. The Federal Circuit disallowed the lost-profits award as impermissibly extraterritorial. The Court reversed 7-2, and the methodology is the notable part: it resolved the case at step one of the presumption-against-extraterritoriality framework, holding that the focus of § 284 is "the infringement," and where the infringement is a domestic act of supplying, an award of lost foreign profits is a permissible domestic application of the damages provision. The Court expressly declined to address proximate cause, which is where these disputes are now litigated.

Impression Products, Inc. v. Lexmark International, Inc. (2017) held patent exhaustion absolute on first authorized sale. Lexmark sold toner cartridges under a single-use/no-resale restriction and sued a remanufacturer. The Court held 8-0 on the domestic question that "a patentee's decision to sell a product exhausts all of its patent rights in that item, regardless of any restrictions the patentee purports to impose." Post-sale restrictions may be enforceable in contract but cannot be enforced through an infringement action, because "the sale transfers the right to use, sell, or import." On the foreign question the Court held 7-1, following Kirtsaeng's copyright logic, that "an authorized sale outside the United States, just as one within the United States, exhausts all rights under the Patent Act." The consequence for case structure: a gray-market or aftermarket dispute that was pleaded as infringement before 2017 is now a breach-of-contract case, which frequently means state court, diversity jurisdiction, and no Federal Circuit appeal.

Minerva Surgical, Inc. v. Hologic, Inc. (2021) preserved but confined assignor estoppel, the doctrine barring an inventor who assigned a patent from later challenging its validity. Minerva's founder had assigned the application; Minerva argued the doctrine should be abolished, Hologic that it should apply categorically. The Court did neither, holding 5-4 that assignor estoppel is valid "but only when the assignor's claim of invalidity contradicts explicit or implicit representations he made in assigning the patent." The opinion gives three examples of where it does not apply: an employee's assignment of future inventions before any claims exist; a later change in law that renders the warranty irrelevant; and — most importantly in practice — a post-assignment amendment that materially broadens the claims beyond what the assignor warranted. That last exception is the one litigated, because continuation practice routinely broadens claims after assignment.

Amgen Inc. v. Sanofi (2023) restored enablement as a meaningful constraint on genus claims. Amgen's claims covered the entire genus of antibodies that bind a specified region of the PCSK9 protein and block its function, defined by function rather than structure, while the specification disclosed the amino-acid sequences of twenty-six examples plus two methods — "roadmapping" and "conservative substitution" — for finding others. The Court held the claims not enabled, unanimously. Section 112(a) requires a specification enabling a skilled artisan "to make and use the entire class of inventions" claimed; the disclosure of a research plan is not enablement, because "a specification may call for a reasonable amount of experimentation to make and use a patented invention," but Amgen's approach left the public "to engage in painstaking experimentation to see what works." The decision has reshaped prosecution and invalidity strategy in biologics and chemistry, where functional genus claims were standard.

Practice checklist

1. Confirm § 1338(a) jurisdiction on the face of the complaint. Plead § 1338 for jurisdiction and §§ 271 and 281 for the claim. If the case is patent-adjacent rather than a patent case, run the four Gunn elements explicitly before filing federally — a hypothetical patent question is not substantial.

1. Run the alternative-theories test before invoking the Federal Circuit. Under Christianson, a claim with more than one supporting theory does not arise under patent law unless patent law is essential to every theory. Audit your own pleading for the non-patent theory that would defeat jurisdiction.

1. Establish each § 1400(b) prong separately and plead it. State of incorporation for the residence prong; acts of infringement and a regular and established place of business for the second. Do not conflate venue with personal jurisdiction — TC Heartland severed them, and Cray rejected contacts-based tests.

1. *Build the Cray record early.* For each alleged place of business, address all three requirements: a physical place, regular and established, and the defendant's. On the third, go to possession and control — ownership or lease, whether employment was conditioned on a district location, advertised addresses, and stored inventory.

1. Remember who carries the burden. Under ZTE, the plaintiff must establish venue, and Federal Circuit law — not regional circuit law — governs the analysis. A defendant's motion can be a short denial supported by a declaration; the plaintiff needs affirmative proof and may need to request venue discovery.

1. Do not rely on infrastructure or remote employees. After Google, a place of business requires the regular physical presence of an employee or agent conducting the defendant's business. Servers in a third party's facility, cloud capacity, and telecommuters' homes ordinarily fail.

1. Treat foreign and domestic defendants differently. A foreign corporation not resident in any district may be sued in any district under § 1391(c)(3); a U.S.-incorporated subsidiary of a foreign parent gets the full benefit of TC Heartland. Check the state of incorporation of every named entity, not the corporate family.

1. Plan the § 1404(a) motion from the outset. Most post-TC Heartland venue fights are convenience fights. Identify named non-party witnesses and their locations, where the accused-product engineers and documents actually are, and the absence of any connection to the forum. Preserve mandamus by making the record complete in the district court.

1. Check § 299 on any multi-defendant complaint. Unrelated accused infringers cannot be joined or tried together based solely on infringing the same patent. Joinder is waivable, and pretrial coordination is still available, but a single complaint against unrelated defendants is defective.

1. *For a declaratory action, plead the controversy under MedImmune.* A licensee in good standing may sue without breaching. Identify the specific accused product and the patentee's specific conduct — demand letters, claim charts, licensing overtures naming the product. Expect a discretionary-declination argument if a mirror-image infringement suit is pending elsewhere.

1. *If you are mooting an invalidity counterclaim, draft the covenant to Already standards.* Unconditional, irrevocable, past and future, reaching distributors and customers, and covering colorable imitations. Anything narrower leaves the counterclaim alive and the covenant in evidence.

1. Coordinate the district court case with PTAB timing. Track the § 315(b) one-year clock from service of the complaint, understand that after Thryv a time-bar ruling is unappealable, and remember that after SAS institution is all-or-nothing with the estoppel consequences that follow. Consider Director review before the § 319 appeal.

1. Preserve the standard of review. Under Teva and Dennison, secure explicit factual findings supported by extrinsic evidence if you win claim construction below; frame the dispute as purely intrinsic if you intend to appeal. Under Highmark and Halo, fee and enhancement rulings get abuse-of-discretion review, so the district court record is the record that matters.

Frequently asked questions

Can a patent infringement claim ever be litigated in state court?

No. Section 1338(a) makes federal jurisdiction over claims arising under the patent laws exclusive of the states. But the question that actually arises is narrower, and the answer is often yes for patent-adjacent claims: license and royalty disputes, assignment and ownership questions governed by contract, invention-assignment employment claims, and legal malpractice arising from a lost patent case all belong in state court. Gunn v. Minton is the controlling authority — a state court may decide a hypothetical patent question as an ingredient of a state-law claim without offending § 1338, because such a question is not substantial to the federal system and the federal-state balance favors state resolution of professional-malpractice claims.

*Where can a domestic corporation be sued for patent infringement after TC Heartland?*

Only in (1) its state of incorporation, under the residence prong of § 1400(b), or (2) a district where it has committed acts of infringement and has a regular and established place of business. Personal jurisdiction is not enough — that was the VE Holding rule TC Heartland rejected. Note that the residence prong is state-wide: a corporation incorporated in Texas may be sued in any of the four Texas districts, because it "resides" in the state, while a Delaware corporation with no Delaware facility may still be sued in any of Delaware's districts.

What makes a location a "regular and established place of business"?

Three things under In re Cray, all required: a physical place in the district (not a virtual presence or an electronic system); a place that is regular (steady and methodical, not sporadic) and established (settled and not transient); and a place that is the defendant's, judged by possession and control — ownership or lease, employment conditioned on a district location, addresses the defendant advertises as its own, inventory stored for sale. In re Google adds a fourth practical requirement: the regular physical presence of an employee or agent conducting the defendant's business. A telecommuting employee's home is not the defendant's place; self-installed servers in a third party's data center are not a place of business.

Who has to prove venue, and whose law applies?

The plaintiff bears the burden of establishing that venue is proper, and Federal Circuit law — not regional circuit law — governs the § 1400(b) analysis. Both holdings come from In re ZTE (USA) Inc. This is a meaningful allocation: a regional-circuit rule that accepts uncontradicted venue allegations as true does not apply, so a defendant's declaration denying possession and control shifts the practical burden back to the plaintiff to produce evidence.

If venue is proper, is the venue fight over?

Usually not. After TC Heartland redistributed filings to districts where defendants genuinely have offices, most venue disputes are § 1404(a) convenience motions rather than § 1400(b) propriety motions. The convenience analysis is governed by regional circuit law, but the Federal Circuit reviews denials by mandamus with unusual willingness — particularly where the district court undervalued the location of documents and non-party witnesses or overweighted its own docket speed. Build the transfer record as if the court of appeals will read it, because it may.

Does a patent counterclaim send the appeal to the Federal Circuit?

Yes, today. Holmes Group, Inc. v. Vornado held in 2002 that it did not, applying the well-pleaded complaint rule to appellate jurisdiction. Congress abrogated that holding in the 2011 AIA by amending § 1295(a)(1) to reach any civil action in which a party asserted a compulsory counterclaim arising under the patent laws, and by adding § 1454 to permit patent-specific removal by any party. Cite Holmes Group for the well-pleaded complaint principle, never for its result.

How much of what the PTAB does is reviewable?

Very little on the institution side; the final written decision is reviewable under § 319. Cuozzo Speed Technologies v. Lee held § 314(d) bars review of institution decisions where the challenge is closely tied to the institution standard, and Thryv v. Click-To-Call extended that to § 315(b) time-bar rulings. The openings are constitutional challenges and claims that the Board exceeded statutory limits in ways untethered to institution — the route SAS Institute v. Iancu used to hold partial institution unlawful, because the objection went to the Board's obligations once instituted rather than to whether it should institute.

Is inter partes review constitutional, and who may use it?

Oil States Energy Services v. Greene's Energy Group held that IPR violates neither Article III nor the Seventh Amendment, because a patent is a public franchise and its reconsideration by the granting agency falls within the public-rights exception. The Court reserved retroactivity, due process, and takings challenges. As to who may petition, Return Mail v. United States Postal Service held the federal government is not a "person" eligible to file, so federal agencies litigate patent exposure in the Court of Federal Claims under 28 U.S.C. § 1498. United States v. Arthrex held the Board's administrative patent judges were unconstitutionally appointed given their unreviewable authority, and cured the defect by making final written decisions subject to Director review — the procedure the PTO has used since 2021.

Why do so many patent standards look like ordinary civil standards now?

Because the Supreme Court replaced Federal Circuit-specific tests one at a time. Octane Fitness and Highmark did it to exceptional-case fees, Halo to enhanced damages, Nautilus to indefiniteness, SCA Hygiene to laches, and Teva — following Dennison Manufacturing v. Panduit twenty-nine years earlier — to the standard of review for claim construction. In each case the displaced rule had the same shape: a rigid threshold, an elevated burden of proof, and de novo appellate review. The replacement in each case was text-driven meaning, totality-of-the-circumstances discretion, a preponderance standard, and deference to the trial court.

This guide covers the law as of September 2026. Every case and statute above links to its full text on LawDiver.