Avoiding Inequitable Conduct at the USPTO: A Duty-of-Candor Checklist

Intellectual Property > Patent | Published 24 April 2026 - Updated 11 May 2026 | UpLaw - uplaw.us

A patent earned through years of research and millions in spend can be rendered worthless not by a better invention or a stronger prior-art reference, but by a single withheld email, a verb in the wrong tense, or an inconsistent sentence filed with another agency. This checklist builds the compliance discipline that keeps candor the default across an entire portfolio, and it explains the doctrine closely enough that you understand why each step matters.

Keywords: inequitable conduct, duty of candor, 37 C.F.R. § 1.56, Therasense, but-for materiality, specific intent to deceive, information disclosure statement, 37 C.F.R. §§ 1.97-1.98, egregious misconduct, infectious unenforceability, supplemental examination, 35 U.S.C. § 257, Exergen, Rule 9(b), prophetic examples, foreign references, Global Dossier, QPIDS

Inequitable conduct is the rare patent defense that does not attack the invention. It attacks the applicant's honesty. A challenger who proves it does not narrow the patent or invalidate a claim or two; the challenger renders the whole patent unenforceable against everyone, forever, no matter how novel the invention or how airtight the claims. The Federal Circuit, sitting en banc, has called the doctrine the "atomic bomb" of patent law precisely because its blast radius bears no relation to the size of the underlying sin. Therasense, Inc. v. Becton, Dickinson & Co., 649 F.3d 1276, 1288 (Fed. Cir. 2011) (en banc). One inventor's decision to bury one failed experiment can level a portfolio.

The doctrine descends from the equitable "unclean hands" cases the Supreme Court decided in the 1930s and 1940s, when it refused to lend the machinery of the courts to patentees who had lied their way to a monopoly. See Keystone Driller Co. v. General Excavator Co., 290 U.S. 240 (1933); Hazel-Atlas Glass Co. v. Hartford-Empire Co., 322 U.S. 238 (1944); Precision Instrument Mfg. Co. v. Automotive Maintenance Machinery Co., 324 U.S. 806 (1945). The modern rule is narrower and more disciplined than the sprawling doctrine that dominated patent litigation before 2011, but it is no less lethal when it lands. It also spills past the courtroom: a patent procured by fraud on the USPTO can support an antitrust counterclaim under Walker Process Equipment, Inc. v. Food Machinery & Chemical Corp., 382 U.S. 172 (1965), an exceptional-case fee award, and disciplinary referral of the practitioner. The good news for anyone building or defending a patent portfolio is that inequitable conduct is almost entirely preventable. It is a conduct problem, not a luck problem, and conduct can be systematized.

This checklist does that. It is organized as ten sequential phases that track the life of an application from the first inventor interview through litigation, plus a set of worked failure modes and a quick-reference table of authorities. Each phase pairs a short block of checkbox items you can lift directly into a matter file with the doctrinal reasoning that explains why the item exists. Read alongside the doctrinal companion, Inequitable Conduct and the Duty of Candor: Patent Law's Atomic Bomb, and the evidence-side guide, Proving Inequitable Conduct in Patent Prosecution: A Comprehensive Guide, it gives a prosecutor the offense (a clean record) and a litigator the defense (a record that survives a Rule 9(b) motion).

How to use this checklist

Treat the phases as cumulative, not optional. Compliance is not a single filing; it is a habit maintained by everyone who touches an application, from the day of conception until the day the patent issues. Assign an owner to each phase, and dock the work to your prosecution calendar so that the candor steps travel with the substantive ones. The single highest-leverage move in the entire document is the cross-routing database in Phase 4, because it defends against both ordinary human forgetfulness and the doctrine's most feared multiplier, infectious unenforceability. If you build nothing else, build that.

A word on scope. This checklist covers the duty of candor in patent prosecution and ex parte reexamination under 37 C.F.R. §§ 1.56 and 1.555. It does not address the separate and differently worded duty of candor before the Patent Trial and Appeal Board under 37 C.F.R. § 42.11, which governs inter partes review and post-grant review practice. The two duties overlap in spirit but not in detail.

Phase 1 — Master the standard you are defending against

You cannot run a compliance program against a doctrine you have not internalized. Since 2011, the controlling framework has been Therasense, and its central move was to raise the bar in two directions at once: materiality and intent must each be proved, and each must be proved to a high standard.

But-for materiality

Before Therasense, a reference was "material" under a loose, USPTO-flavored standard whenever a reasonable examiner might have wanted to see it. The en banc court scrapped that in favor of a stringent test borrowed from harmless-error analysis: information is material only if the USPTO would not have allowed the claim had it known the truth. Therasense, 649 F.3d at 1291. In practice the court asks the "would-the-claim-have-issued" question by giving claims their broadest reasonable construction and applying a preponderance standard, the way an examiner would. If the claim still would have issued, the withheld information was not but-for material, and the inquiry usually ends there.

The practical upshot is asymmetric. A challenger's task got harder, but a prosecutor's job did not get easier, because you almost never know in advance which references a future court will deem but-for material. The prudent assumption is that anything genuinely relevant to patentability is material, and the compliance program disclosures accordingly. Note too what the USPTO's own regulation still says: the agency's duty-of-disclosure rule, 37 C.F.R. § 1.56(b), defines materiality more broadly than Therasense does, and the USPTO has never finished harmonizing its rule with the case law despite proposing to do so. See 81 Fed. Reg. 74,987 (Oct. 28, 2016); 76 Fed. Reg. 43,631 (July 21, 2011). Prosecutors therefore live under two materiality standards at once: the broad regulatory duty that governs their conduct before the Office, and the narrow Therasense standard that a court will apply years later. Compliance means satisfying the broader of the two.

Specific intent to deceive

Materiality is only half the case. The challenger must independently prove that the applicant acted with the specific intent to deceive the Office. Therasense holds that the accused infringer must show the applicant "knew of the reference, knew that it was material, and made a deliberate decision to withhold it." 649 F.3d at 1290. And the evidentiary bar is exacting: deceptive intent must be "the single most reasonable inference able to be drawn from the evidence." Id. If the record supports multiple reasonable inferences, one of which is innocent, the challenger loses. Negligence will not do; gross negligence will not do. See Kingsdown Medical Consultants, Ltd. v. Hollister Inc., 863 F.2d 867, 876 (Fed. Cir. 1988) (en banc) (gross negligence does not by itself justify an inference of intent to deceive).

The Federal Circuit has enforced this rigor. In 1st Media, LLC v. Electronic Arts, Inc., 694 F.3d 1367 (Fed. Cir. 2012), the court reversed an inequitable-conduct finding because "should have known" is not "knew and chose to deceive." At the same time, intent can be inferred from circumstantial evidence, and a sufficiently damning pattern still sinks patents. In Aventis Pharma S.A. v. Hospira, Inc., 675 F.3d 1324 (Fed. Cir. 2012), and Apotex Inc. v. UCB, Inc., 763 F.3d 1354 (Fed. Cir. 2014), the court affirmed unenforceability where inventors made affirmative misrepresentations that no innocent explanation could account for. And in American Calcar, Inc. v. American Honda Motor Co., 768 F.3d 1185 (Fed. Cir. 2014), the court applied the tightened standard on remand and still found intent, proving that Therasense raised the bar without abolishing the defense.

The death of the sliding scale

The most important structural change Therasense made is easy to miss: it killed the "sliding scale." Under the old regime, a strong showing of materiality could compensate for a weak showing of intent, and vice versa. No longer. Materiality and intent are now separate elements, each proved on its own. Therasense, 649 F.3d at 1290. A challenger cannot leverage a devastating undisclosed reference into a finding of intent; the intent case must stand on its own feet. For the compliance program this cuts both ways. It means a genuine, documented good-faith judgment call about a borderline reference is a real defense even if a court later deems the reference material. It also means you cannot buy your way out of a bad intent record with a weak materiality argument. The two questions are answered independently, and you must win both.

The egregious-misconduct exception

Therasense carved one exception into the but-for materiality requirement. Affirmative acts of egregious misconduct, most prominently the filing of an unmistakably false affidavit or declaration, are material as a matter of law, without any but-for inquiry. 649 F.3d at 1292. The court reasoned that a fabricated declaration works a fraud on the Office regardless of whether the lie changed the outcome. The exception is narrow and deliberately so: mere nondisclosure of a reference never qualifies, and the court warned against letting the exception swallow the rule. But when it applies, it is brutal, because the challenger skips the hardest part of its case. The lesson for the program is blunt: never let a false declaration enter the file, and if one already has, correct it loudly (Phase 8). See Intellect Wireless, Inc. v. HTC Corp., 732 F.3d 1339 (Fed. Cir. 2013); Transweb, LLC v. 3M Innovative Props. Co., 812 F.3d 1295, 1304 (Fed. Cir. 2016).

Infectious unenforceability

Finally, understand the doctrine's terrifying reach across a family. Under Consolidated Aluminum Corp. v. Foseco International Ltd., 910 F.2d 804 (Fed. Cir. 1990), inequitable conduct committed while prosecuting one patent can render a related patent unenforceable, even where nothing improper happened during the related patent's own prosecution, so long as the two bear an "immediate and necessary relation" to the misconduct. A tainted parent can poison its continuations. This is why candor cannot be run application by application; it must be run portfolio-wide. The cross-routing discipline in Phase 4 is the direct answer to Consolidated Aluminum.

Phase 1 checklist

• [ ] Confirm the team understands the duty of candor under 37 C.F.R. § 1.56: everyone substantively involved must disclose information material to patentability that they know of.
• [ ] Understand the Therasense test: but-for materiality and specific intent to deceive, each by clear and convincing evidence, with deceptive intent the single most reasonable inference.
• [ ] Note that materiality and intent are independent; the sliding scale is dead.
• [ ] Note the narrow egregious-misconduct exception: an unmistakably false affidavit or declaration is material per se.
• [ ] Recall infectious unenforceability: misconduct in one patent can spread to related patents (Consolidated Aluminum).
• [ ] Remember the USPTO's regulatory materiality standard (§ 1.56(b)) is broader than Therasense; comply with the broader one.

Phase 2 — Identify everyone who owes the duty

The duty of candor does not attach to a law firm or a corporation as an abstraction. It attaches to individuals, and the roster is wider than most inventors assume. Under 37 C.F.R. § 1.56(c), the duty runs individually to each attorney or patent agent prosecuting the application, to each named inventor, and to every other person who is substantively involved in the preparation or prosecution of the application and associated with the inventor, the applicant, or the assignee. That last category is the sleeper. It reaches in-house counsel, the scientist who ran the confirmatory assay, the regulatory-affairs manager who filed a parallel submission with the FDA, the licensing executive who knows about a prior offer for sale, and sometimes a company officer. If a person knows material information and is substantively involved, that person owes the duty, whether or not they ever speak to the examiner.

Two mechanics matter. First, each named inventor must acknowledge the duty in the inventor's oath or declaration under 37 C.F.R. § 1.63(c); this is not boilerplate, it is a sworn undertaking. Second, an individual discharges the duty by disclosing the material information to the prosecuting attorney or agent. 37 C.F.R. § 1.56(d). That single rule is the architecture of every good compliance program: you create reliable channels for material information to reach counsel, and counsel decides what to file. The inventor's job is to tell, not to judge.

Phase 2 checklist

• [ ] Map every named inventor and confirm each signs the § 1.63(c) declaration understanding it as a real obligation.
• [ ] Map each prosecuting attorney and patent agent.
• [ ] Map every other substantively involved person: in-house counsel, scientific and technical staff, regulatory personnel, licensing managers, and officers.
• [ ] Confirm everyone understands the duty runs throughout pendency, up to grant, and that disclosing to the prosecuting attorney discharges the individual's duty (§ 1.56(d)).

Phase 3 — Build a candor culture through education

Educated clients comply; confused clients withhold. The single most cost-effective control in the entire program is a plain-language briefing given to every inventor and every substantively involved employee at the start of a matter, and repeated. It should explain, without jargon, four things: what must be disclosed, who is covered, that the duty continues all the way to grant, and that a violation can cost the entire patent and its whole family. Frame the § 1.63(c) acknowledgment as the sworn statement it is. And, as counsel, do the small ritual that matters more than it looks: repeat a short candor reminder in every substantive written communication during prosecution. A one-sentence standing reminder in each status letter builds a contemporaneous record that the duty was live in everyone's mind, which is exactly the kind of evidence that defeats an inference of deliberate concealment.

A few points inventors find genuinely clarifying, because they cut against intuition. Disclosing a reference is not an admission that it is material; the rules say so expressly. 37 C.F.R. § 1.97(h). There is no duty to go hunting for prior art, although a search is almost always advisable and often required for other reasons. Material information reaches far beyond novelty and obviousness; it can bear on enablement, written description, inventorship, ownership, and inconsistent statements. And "material" is decided by counsel, not by the inventor at the bench. Tell the inventor to over-report and let counsel triage. The upstream discipline of a good invention disclosure is where this culture is seeded; an inventor trained to document conception, alternatives, and known prior art is an inventor already practicing candor. For the strategic version of the same task, see Drafting an Invention Disclosure That Maximizes Patent Scope.

Phase 3 checklist

• [ ] Give every inventor and relevant employee a plain-language briefing on the duty, its scope, its duration, and the stakes.
• [ ] Present the § 1.63(c) acknowledgment as a real, sworn obligation.
• [ ] Repeat a short candor reminder in every substantive client communication during prosecution.
• [ ] Explain the counterintuitive points: disclosure is not an admission of materiality (§ 1.97(h)); materiality is counsel's call; over-report.

Phase 4 — Gather and cross-route material information

This is the operational heart of the program. Everything else is doctrine; this is plumbing. The goal is a system that reliably surfaces material information from every corner of the enterprise and routes any item found in one matter to every related matter, automatically.

Start by casting a wide net at the source. Instruct inventors to surface everything potentially relevant: prior products and publications, public uses and offers for sale, all testing including the experiments that failed, competitor activity, and any related litigation or regulatory filings. Then, as counsel, identify the material categories the case law and the MPEP flag: prior-art patents, products, publications, and activities under 35 U.S.C. § 102; potential double-patenting references; enablement and written-description data under § 112; inventorship and ownership facts, see Perceptive Biosystems, Inc. v. Pharmacia Biotech, Inc., 225 F.3d 1315 (Fed. Cir. 2000); office actions, search reports, and cited art from related U.S. and foreign counterpart applications; conflicting submissions to other agencies such as the FDA; and any statement inconsistent with a position taken before the Office. On the double-patenting front, families that pursue both a device and its ornamental design create their own cross-citation obligations; see Combining Design and Utility Patents: Concurrent Protection and Double-Patenting Pitfalls.

Now build the database. The workable form is a living spreadsheet with documents in the rows and related applications in the columns, so a glance tells you whether item X has been cited in application Y. Designate a group of interrelated families to be cross-referenced. Monitor incoming correspondence, office actions, search reports, litigation productions, for new items. Update the matrix when items arrive and again when IDSs are filed. This is not busywork; it is the direct antidote to infectious unenforceability, because it guarantees that a reference cited by a German examiner in a counterpart lands in the U.S. file of every sibling application before anyone can accuse you of hiding it. Many portfolios use a commercial docketing service for this; the tool matters less than the discipline.

Two force multipliers. First, the USPTO's Global Dossier consolidates family-member file histories from participating offices, so foreign search reports and cited references can be pulled into the U.S. record rather than reconstructed by hand. Use it. Second, hold the line on the division of labor: inventors list what they know; counsel decides materiality. An inventor who prejudges a reference as immaterial and drops it is exactly how patents die.

Worked example — the failed batch. Helix Genomics files a utility application on an enzyme variant, reciting data from confirmatory runs Batch 3 through Batch 9. Batch 7 failed spectacularly; the variant lost all activity under the very conditions the claims recite. A junior scientist, embarrassed, leaves Batch 7 out of the summary handed to counsel and mentions it to no one. Two years later, in litigation, the defendant's expert finds Batch 7 in a lab notebook produced in discovery and argues the omission was material to enablement and made with intent to deceive. Whether the omission is ultimately but-for material is a fight; but the fight itself, and the discovery of a hidden failed experiment, is precisely what a disciplined intake process (Phase 3) and a complete data hand-off (this phase) exist to prevent. Failed experiments are not embarrassments to be buried. They are the data a court will use against you if you hide them, and harmless disclosures if you don't.

Phase 4 checklist

• [ ] Instruct inventors to surface all potentially relevant information, expressly including failed experiments, public uses, offers for sale, and competitor activity.
• [ ] Identify the material categories: § 102 prior art, double-patenting references, § 112 data, inventorship and ownership facts, related-application and foreign-counterpart office actions and search reports, conflicting other-agency submissions, and inconsistent statements.
• [ ] Build and maintain a centralized cross-routing database that pushes any item found in one matter to every related matter.
• [ ] Use Global Dossier to pull foreign family office actions and cited references into the U.S. record.
• [ ] Enforce the division of labor: inventors list what they know; counsel decides materiality (§ 1.97(h)).

Phase 5 — Draft candor-compliant specifications and declarations

The specification and the declarations are where affirmative misstatements live, and affirmative misstatements are more dangerous than omissions because they invite the egregious-misconduct exception and are harder to explain away as an honest oversight.

Do not mischaracterize the prior art. The specification does not have to discuss cited references at all; a bland application that simply claims the invention is safer than one that editorializes about what came before and gets it wrong. Where you must describe the art, describe it accurately.

Handle prophetic examples with religious care. Patent practice permits "prophetic" or "paper" examples, experiments that describe what an inventor predicts will happen but has not actually run, and the MPEP even prescribes the grammatical tell: prophetic examples are written in the present or future tense, while examples of work actually performed are written in the past tense. MPEP § 608.01(p). Cross that line, present a predicted result as a completed one, and you have manufactured a false statement of fact. The tense of a single verb has driven the analysis. See Purdue Pharma L.P. v. Endo Pharmaceuticals Inc., 438 F.3d 1123, 1131 (Fed. Cir. 2006); Novo Nordisk Pharmaceuticals, Inc. v. Bio-Technology General Corp., 424 F.3d 1347, 1362 (Fed. Cir. 2005).

Worked example — the tense trap. Nimbus Materials drafts an application on a coating and includes "Example 4," which reads, "The substrate was coated and exhibited a 40% reduction in wear." No such test was run; the 40% figure is a model prediction. Written in the past tense, Example 4 now reads as a completed experiment reporting real data. If wear performance later becomes the pivot of a validity fight, the accused infringer will argue that Example 4 is a fabricated result, and the egregious-misconduct exception looms because it is an affirmative misstatement rather than a mere omission. Had the drafter written "The substrate is coated and exhibits," the same content would have been an unremarkable prophetic example. One verb, two very different records.

Do not cherry-pick data. Selectively reporting the runs that worked while suppressing the ones that did not can distort the objective meaning of the disclosure and the arguments built on it, and it is a recognized species of candor violation. If a data set is mixed, present it as mixed.

Scrutinize every declaration and affidavit as if a hostile expert will parse it word by word, because one will. Each factual assertion must be literally true, and, just as important, the declaration must not create a false impression by careful omission. A declaration that is technically true but engineered to mislead is treated as a false declaration. Intellect Wireless, Inc. v. HTC Corp., 732 F.3d 1339 (Fed. Cir. 2013). In Intellect Wireless, an inventor's declaration about "actual reduction to practice" of a wireless picture-phone was, at best, misleading, and the belated, hedged "correction" did not cure it; the Federal Circuit affirmed unenforceability.

Reconcile your statements across agencies. Telling the USPTO that a formulation range is "critical" and unexpectedly effective while telling the FDA that the same range is old and well known is the kind of two-faced record that ends patents. That is essentially what happened in Belcher Pharmaceuticals, LLC v. Hospira, Inc., 11 F.4th 1345 (Fed. Cir. 2021), where the patentee's inconsistent characterizations of a pH range, critical to the Patent Office, unremarkable to the FDA, produced an affirmed finding of inequitable conduct. The USPTO has since emphasized that the duty of candor reaches information from proceedings at other agencies. See 87 Fed. Reg. 45,764 (July 29, 2022).

Worked example — two faces of a range. Meridian Pharma prosecutes a claim to an injectable stabilized within a pH of 2.8 to 3.3, arguing to the examiner that the narrow range is the unexpected key to shelf stability. In a contemporaneous FDA submission, Meridian's regulatory team writes that the same pH range is conventional and long used in the field, to smooth its approval path. Both statements sit in files that a litigant can subpoena. Read together, they suggest that Meridian told the Office the range was novel while privately treating it as old news, the Belcher pattern. The cross-routing database in Phase 4 is supposed to catch exactly this, by making the regulatory submission visible to prosecution counsel before the inconsistent argument is ever filed.

Get inventorship right. A deliberate misnaming or omission of an inventor can itself be inequitable conduct; honest errors, by contrast, are correctable under 35 U.S.C. § 256 without penalty. Inventorship is increasingly fraught where AI tools contribute to conception; see Who Owns AI-Generated Inventions? Machine Creativity and Patent Ownership. Chain-of-title and assignment discipline sits adjacent to inventorship candor; see Employee Invention Assignment Agreements: Securing Clean Title Across Jurisdictions.

Phase 5 checklist

• [ ] Do not mischaracterize the prior art; you need not describe cited references at all.
• [ ] Keep prophetic examples in present or future tense; never present a predicted experiment as completed (MPEP § 608.01(p); Purdue Pharma; Novo Nordisk).
• [ ] Do not cherry-pick favorable data while suppressing unfavorable results.
• [ ] Scrutinize every declaration; each assertion must be literally true and must not mislead by omission (Intellect Wireless).
• [ ] Reconcile statements across agencies; do not tell the USPTO one thing and the FDA the opposite (Belcher).
• [ ] Get inventorship right; deceptive naming can be inequitable conduct, while innocent errors are correctable under § 256.

Phase 6 — File timely, complete Information Disclosure Statements

The Information Disclosure Statement is the mechanism by which material information is formally placed before the examiner. Nothing else counts. Citing a reference in the specification, or mentioning it in an office-action response, does not discharge the duty; the USPTO does not formally consider information disclosed by other means. MPEP § 609.05(c). The IDS rules live at 37 C.F.R. §§ 1.97 and 1.98, and the single most important operational fact about them is that the requirements get progressively heavier the later you file. File early and the path is clean; file late and you must certify, pay, or both.

The IDS timing windows work as follows:

Filing window	Requirement	Practical note
Within 3 months of filing, or before the first office action on the merits	No fee, no certification	The cleanest path. Front-load everything you have. 37 C.F.R. § 1.97(b).
After the first office action, before final rejection or allowance	Fee or § 1.97(e) certification	Pay the fee to avoid certifying. 37 C.F.R. § 1.97(c).
After final rejection or notice of allowance	Fee and § 1.97(e) certification	If you cannot truthfully certify, do not file; file an RCE with the IDS instead. 37 C.F.R. § 1.97(d).
After issue-fee payment	QPIDS, or withdraw-from-issue plus RCE, or a continuing application, or supplemental examination	No 3-month grace period applies. 37 C.F.R. §§ 1.114, 1.313(c)(2).

The § 1.97(e) certification is a sworn statement, for each item, that the item was either first cited in a foreign counterpart within the last three months, or was not known to the applicant more than three months before the filing. That certification about foreign counterparts is not softened by any lack of personal knowledge; you cannot hedge it. MPEP § 609.04(b). The rule that follows is one of the most important in this entire checklist and is worth stating flatly: if counsel cannot truthfully make the certification, counsel must not file the IDS. Filing a false certification to squeeze a late reference into an allowed case converts a curable timing problem into an affirmative misstatement, which is far worse. The right move is a Request for Continued Examination with the IDS attached, which reopens prosecution and lets the examiner consider the reference cleanly. 37 C.F.R. § 1.114.

Content matters as much as timing. Under 37 C.F.R. § 1.98, the IDS must list each item, provide legible copies of everything except U.S. patents and U.S. published applications, and, for foreign-language references, provide any existing English translation or a concise explanation of relevance. Do not skimp on the copies or the translations; an incomplete IDS may not be considered at all.

Then close the loop. After a notice of allowance, review the file to confirm the examiner initialed each reference as "considered" and signed each IDS. MPEP § 609.05. An uninitialed reference was not formally placed before the Office, which means the protective value of disclosure, an examiner who considered the art is much harder to accuse of being deceived, evaporates. The review after allowance should confirm three things: that counsel disclosed all known material information, that the examiner initialed every citation, and that no uncorrected misstatement remains in the record.

Finally, file supplemental IDSs promptly whenever new material information surfaces, after an office action cites new art, after a foreign counterpart is examined, or after litigation turns up a reference you did not have. The duty is continuous, and the cross-routing database of Phase 4 is what makes continuous compliance mechanical rather than heroic. This work dovetails with the broader task of responding to a patent office action; every response is also an occasion to check whether a supplemental IDS is due. For the full prosecution arc, see the patent prosecution roadmap.

Phase 6 checklist

• [ ] File the initial IDS in the early window (within 3 months of filing or before the first office action) for no fee and no certification.
• [ ] After the first office action, file with the fee or § 1.97(e) certification.
• [ ] After final rejection or allowance, file with fee and certification; if you cannot truthfully certify, file an RCE with the IDS instead.
• [ ] After issue-fee payment, use QPIDS, withdrawal-from-issue plus RCE, a continuing application, or supplemental examination.
• [ ] Meet § 1.98 content rules: list each item, supply copies of non-U.S.-patent references, and provide relevance explanations and translations for foreign-language references.
• [ ] Verify the examiner initialed each reference as "considered" and signed each IDS (MPEP § 609.05).
• [ ] File supplemental IDSs promptly whenever new material information surfaces.

Phase 7 — Avoid the IDS traps

An IDS filed the wrong way can be worse than none at all, because it can look like an attempt to comply while actually concealing.

Do not bury the critical reference. There is a tension in candor practice between disclosing generously and disclosing usefully. Dump two hundred marginal references on an examiner alongside the one that actually reads on the claims, and a later challenger will argue you deliberately hid the important art in a haystack. The answer is not to disclose less; it is to disclose everything material and, where a reference is genuinely central, to flag it, in a relevance explanation, an examiner interview, or the response, so no one can say you tried to sneak it past. Generous disclosure and honest signposting are not in conflict.

Handle foreign-language references honestly. The trap here is selective translation, submitting the reference and a translation of only the passages that help you while quietly omitting the passages that hurt. That is a candor violation dressed up as compliance. Semiconductor Energy Laboratory Co. v. Samsung Electronics Co., 204 F.3d 1368 (Fed. Cir. 2000). Submit the full reference, submit whatever translation you have, and expressly flag the translation's limits. If you have only a partial translation, say so.

Do not let patent-term-adjustment optimization tempt you into delaying or omitting a disclosure. It is true that a late IDS can cost a few days of patent term under the PTA rules, 37 C.F.R. § 1.704(c). It is also true that no number of extra term days is worth the risk of unenforceability. Never let the tail of PTA arithmetic wag the dog of candor.

Phase 7 checklist

• [ ] Do not bury critical references in a haystack of marginal ones; disclose everything material but flag the genuinely central references.
• [ ] For foreign-language references, submit the reference, submit the translation you have, and flag its limits; never translate only the favorable passages (Semiconductor Energy Laboratory).
• [ ] Do not delay or omit a disclosure to optimize patent-term adjustment.

Phase 8 — Correct errors expressly and unmistakably

If you discover that the record contains a misstatement or a misleading impression, in the specification, an argument, or a declaration, the cure is an express, conspicuous correction that identifies both the error and the corrected information. A quiet walk-back does not work. This is the specific holding that trapped the patentee in Intellect Wireless: the inventor's supplemental declaration hedged and softened the original false statement without ever squarely retracting it, and the court held the "correction" insufficient to purge the taint. 732 F.3d at 1345; see MPEP § 2011. The mental model is a retraction printed in the same size type as the original, not a footnote hoping no one reads it. Name the error. State the truth. File it where the examiner will see it. Ambiguity in a correction is read against the patentee, because a half-hearted correction looks like an attempt to have it both ways.

Phase 8 checklist

• [ ] On finding any misstatement or misleading impression, make an express correction that identifies both the error and the corrected information (Intellect Wireless; MPEP § 2011).
• [ ] Do not attempt a quiet or ambiguous walk-back; it will not cure the taint.

Phase 9 — Cure post-grant problems with supplemental examination

Sometimes a candor gap surfaces only after the patent has issued, when the asset is valuable and in force. The America Invents Act created a purpose-built tool for exactly this situation: supplemental examination under 35 U.S.C. § 257. It lets a patent owner ask the USPTO to consider, reconsider, or correct information believed relevant to the patent, and, crucially, it reaches any question of patentability, not just patents and printed publications. That breadth is what distinguishes it from ordinary ex parte reexamination.

Its signature feature is immunity. If the USPTO conducts supplemental examination and any resulting reexamination on the information, the patent cannot later be held unenforceable for inequitable conduct on the basis of that information. 35 U.S.C. § 257(c)(1). In effect, supplemental examination lets a patentee launder a candor problem through the Office and emerge protected, on the specific information considered.

But the timing rules are unforgiving, and they punish procrastination. The immunity does not apply if the allegation of inequitable conduct was pled with particularity, or asserted with particularity in a notice under the Hatch-Waxman framework, before the supplemental-examination request was filed. 35 U.S.C. § 257(c)(2)(A). And in a patent-enforcement action, the immunity is lost unless the supplemental examination, and any resulting reexamination, concluded before the enforcement suit was brought. § 257(c)(2)(B). The practical rule writes itself: the protection is available only to the patentee who moves before trouble is pled. Once you are in the crosshairs, once a defendant has pled inequitable conduct with Rule 9(b) particularity, the shield is gone. Supplemental examination is medicine you take before you are sick.

Two more cautions. First, weigh supplemental examination against its cousins. A reissue under 35 U.S.C. § 251 can fix certain errors but carries none of § 257's immunity, and patent-owner-initiated ex parte reexamination likewise lacks the shield. Choose the tool that fits the defect. Second, supplemental examination is not a criminal or antitrust shelter. If the USPTO becomes aware of a material fraud in the course of supplemental examination, it may refer the matter to the Attorney General. § 257(e). And nothing in § 257 immunizes a patentee against Walker Process antitrust exposure for enforcing a fraudulently procured patent. The immunity is real but bounded; do not oversell it to a client.

Phase 9 checklist

• [ ] For a candor gap in a valuable, in-force patent, evaluate supplemental examination under 35 U.S.C. § 257; it reaches any patentability question and can immunize the patent on the information considered (§ 257(c)(1)).
• [ ] Act before any inequitable-conduct allegation is pled with particularity; the immunity is forfeited once you are in the crosshairs (§ 257(c)(2)).
• [ ] Compare with reissue (§ 251) and patent-owner ex parte reexamination, neither of which carries § 257's immunity.
• [ ] Remember the USPTO may refer material fraud to the Attorney General (§ 257(e)); supplemental examination is no shield against criminal fraud or Walker Process antitrust liability.

Phase 10 — Defend a pleaded inequitable-conduct charge

If, despite everything, an accused infringer pleads inequitable conduct, the defense begins with the pleading itself. Inequitable conduct sounds in fraud, and Federal Rule of Civil Procedure 9(b) therefore requires it to be pled with particularity. Under Exergen Corp. v. Wal-Mart Stores, Inc., 575 F.3d 1312, 1328-29 (Fed. Cir. 2009), a pleading must identify the specific who, what, when, where, and how of the material misrepresentation or omission, and must allege facts from which a court can reasonably infer both that a specific individual knew of the material information and that the individual withheld or misrepresented it with a specific intent to deceive. Conclusory recitals that "the applicant" hid "material prior art" fail. A well-aimed motion to dismiss or to strike the defense removes the cloud early and cheaply, and it forces the challenger to show its cards before discovery.

If the defense survives the pleadings, the prosecution record becomes the battleground, which is why the earlier phases matter so much in hindsight. A clean, complete file history, with early IDSs, examiner initials, contemporaneous candor reminders, and documented good-faith judgment calls, is the single best defense, because it starves the challenger's intent case. Remember that under Therasense intent must be the single most reasonable inference; every documented good-faith explanation in the file is another reasonable inference that defeats that standard.

Manage privilege deliberately. Asserting that counsel made a good-faith judgment call in reliance on advice can waive privilege over the very prosecution communications you would rather keep. Decide early whether you will rely on advice of counsel, and understand the waiver that follows; a disciplined privilege log and a clear-eyed waiver analysis belong in the litigation plan from day one.

Finally, mind your conduct in the litigation itself. Candor and clean-hands obligations do not end when the patent issues; they follow the patentee into enforcement. In Regeneron Pharmaceuticals, Inc. v. Merus N.V., 864 F.3d 1343 (Fed. Cir. 2017), the district court, faced with a pattern of litigation misconduct, drew an adverse inference of the very deceptive intent that had occurred during prosecution, and the Federal Circuit affirmed. Misbehaving in the case can supply the intent element the challenger otherwise struggled to prove. For the surrounding procedural landscape, see The Patent Infringement Lawsuit From Filing to Appeal: A Comprehensive Guide and the patent litigation roadmap. Inequitable conduct also sits alongside the other equitable defenses a patentee may face, and its practical mechanics interlock with the general anatomy of patent infringement claims and defenses.

Phase 10 checklist

• [ ] Hold a pleaded charge to Rule 9(b): the specific who, what, when, where, and how, with facts supporting a reasonable inference of knowledge and specific intent (Exergen).
• [ ] Preserve a clean prosecution record and contemporaneous documentation of good-faith judgment calls.
• [ ] Manage privilege deliberately; asserting good-faith reliance can waive privilege over prosecution communications.
• [ ] Behave in the litigation; misconduct can support an adverse inference of prosecution-era intent (Regeneron).

Failure modes: six ways good patents die

The doctrine is easier to remember as a set of cautionary tales than as a list of rules. Here are the recurring ways candor programs fail, each mapped to the phase that would have prevented it.

Treating candor application by application. The portfolio-blind prosecutor cites a reference in the parent, forgets it in the continuation, and hands a challenger an infectious-unenforceability theory under Consolidated Aluminum. Prevention lives in the Phase 4 cross-routing database.

Hiding a failed experiment. The buried Batch 7. The instinct to suppress unflattering data is human and fatal, because the suppressed data is exactly what a hostile expert will surface in discovery. Prevention is the complete data hand-off of Phase 4 and the culture of Phase 3.

Filing a false or half-true declaration. The affirmative misstatement is the most dangerous species of candor violation because it opens the door to the egregious-misconduct exception and skips the challenger's hardest work. Prevention is the word-by-word declaration review of Phase 5.

Withholding a related-application rejection or a conflicting agency submission. The examiner in your U.S. case never learns that a foreign counterpart was rejected over art you never cited, or that your regulatory team told the FDA the opposite of what you told the Office. This is the Belcher pattern, and it is caught only by cross-routing everything, including non-patent-office files.

Curing a misstatement quietly. The ambiguous, hedged "correction" that Intellect Wireless condemned. A cure that does not squarely name the error and state the truth is no cure at all. Prevention is the express-correction discipline of Phase 8.

Racing to supplemental examination only after suit. The § 257 immunity is forfeited once inequitable conduct is pled, so the patentee who waits until the defendant files loses the very protection the statute created. Prevention is treating supplemental examination as a proactive tool, per Phase 9.

Quick-reference table of authorities

Regulations and rules. 37 C.F.R. §§ 1.56 (duty of candor), 1.63(c) (inventor's acknowledgment), 1.97 and 1.98 (IDS timing and content), 1.114 (RCE), 1.313(c)(2) (withdrawal from issue), 1.555 (reexamination candor), 1.704 (patent-term adjustment), 11.303 (candor toward the Office); 42.11 (PTAB candor, different regime). 35 U.S.C. §§ 251 (reissue), 256 (inventorship correction), 257 (supplemental examination). Fed. R. Civ. P. 9(b).

Cases. Therasense, Inc. v. Becton, Dickinson & Co., 649 F.3d 1276 (Fed. Cir. 2011) (en banc) (but-for materiality and specific intent, sliding scale abolished, egregious-misconduct exception); Kingsdown Medical Consultants, Ltd. v. Hollister Inc., 863 F.2d 867 (Fed. Cir. 1988) (en banc) (gross negligence insufficient); Exergen Corp. v. Wal-Mart Stores, Inc., 575 F.3d 1312 (Fed. Cir. 2009) (Rule 9(b) pleading); Intellect Wireless, Inc. v. HTC Corp., 732 F.3d 1339 (Fed. Cir. 2013) (false declaration; inadequate cure); Belcher Pharmaceuticals, LLC v. Hospira, Inc., 11 F.4th 1345 (Fed. Cir. 2021) (inconsistent statements across agencies); Regeneron Pharmaceuticals, Inc. v. Merus N.V., 864 F.3d 1343 (Fed. Cir. 2017) (adverse inference from litigation misconduct); Consolidated Aluminum Corp. v. Foseco International Ltd., 910 F.2d 804 (Fed. Cir. 1990) (infectious unenforceability); Semiconductor Energy Laboratory Co. v. Samsung Electronics Co., 204 F.3d 1368 (Fed. Cir. 2000) (selective translation); Purdue Pharma L.P. v. Endo Pharmaceuticals Inc., 438 F.3d 1123 (Fed. Cir. 2006) and Novo Nordisk Pharmaceuticals, Inc. v. Bio-Technology General Corp., 424 F.3d 1347 (Fed. Cir. 2005) (prophetic-example tense); GS Cleantech Corp. v. Adkins Energy LLC, 951 F.3d 1310 (Fed. Cir. 2020), Transweb, LLC v. 3M Innovative Props. Co., 812 F.3d 1295 (Fed. Cir. 2016), Aventis Pharma S.A. v. Hospira, Inc., 675 F.3d 1324 (Fed. Cir. 2012), Apotex Inc. v. UCB, Inc., 763 F.3d 1354 (Fed. Cir. 2014), American Calcar, Inc. v. American Honda Motor Co., 768 F.3d 1185 (Fed. Cir. 2014), and 1st Media, LLC v. Electronic Arts, Inc., 694 F.3d 1367 (Fed. Cir. 2012) (post-Therasense application); Walker Process Equipment, Inc. v. Food Machinery & Chemical Corp., 382 U.S. 172 (1965) (antitrust exposure); Keystone Driller, Hazel-Atlas Glass, and Precision Instrument (equitable roots).

MPEP. §§ 2000.01, 2001.04, 2001.06, 2004, 2005, 2010, 2011, 2016, 608.01(p), 609.04-609.05. USPTO tools. Global Dossier; QPIDS pilot program.

This checklist is general legal information, not legal advice, and does not create an attorney-client relationship. Candor questions turn on specific facts; consult a registered patent practitioner about any particular application or patent.

Related UpLaw Resources

Related Guides and Articles

• Inequitable Conduct and the Duty of Candor: Patent Law's Atomic Bomb
• Proving Inequitable Conduct in Patent Prosecution: A Comprehensive Guide
• Drafting an Invention Disclosure That Maximizes Patent Scope
• Turning Patent Rejections Into Granted Claims: A Prosecution Response Guide
• Combining Design and Utility Patents: Concurrent Protection and Double-Patenting Pitfalls
• Who Owns AI-Generated Inventions? Machine Creativity and Patent Ownership
• How U.S. Patents Work: A Comprehensive Practical Guide
• Equitable Defenses: A Comprehensive Guide to Laches, Acquiescence, Waiver, and Equitable Estoppel
• The Patent Infringement Lawsuit From Filing to Appeal: A Comprehensive Guide
• Patent Infringement, Claims, and Defenses: A Practical Guide for Businesses and Counsel

Related Checklists

• Drafting an Invention Disclosure: A Practical Checklist
• Responding to a Patent Office Action: A Practical Checklist
• Defeating a Section 103 Obviousness Rejection: A Practical Checklist
• Running a Freedom-to-Operate Patent Search: A Practical Checklist
• Building a Privilege Log: A Practical Checklist

Related Toolkits

• Prosecuting a Patent Application: A Roadmap and Research Guide
• Litigating Patent Infringement: A Roadmap and Research Guide