The PTAB is the USPTO's own tribunal, and it sits in two worlds at once.
One half is prosecution. When an examiner finally rejects claims and an applicant appeals, the Board decides. That is the tail end of obtaining a patent.
The other half is adversarial. Inter partes review and post-grant review are trials between two parties over whether an issued patent should have granted.
The second half is why the Board matters commercially, and why it changed patent disputes when it was created in 2012.
Where the PTAB came from
Created by the America Invents Act, operational from September 2012, replacing the Board of Patent Appeals and Interferences.
| Before 2012 | After |
|---|---|
| Board of Patent Appeals and Interferences | Patent Trial and Appeal Board |
| Ex parte appeals | Retained |
| Interference proceedings | Replaced by derivation proceedings |
| Inter partes reexamination | Replaced by inter partes review |
| No §101 post-grant route | Post-grant review added |
Interferences ended with first-inventor-to-file. Determining who invented first stopped mattering once priority turned on filing date, so derivation proceedings — asking whether an applicant derived the invention from someone else — took their place.
The post-grant trials were the substantive addition, and they are what changed patent disputes.
What the Board hears
| Proceeding | Nature | Parties |
|---|---|---|
| Ex parte appeal | Prosecution | Applicant and examiner |
| Inter partes review | Post-grant trial | Petitioner and patent owner |
| Post-grant review | Post-grant trial | Petitioner and patent owner |
| Derivation proceeding | Priority dispute | Two applicants |
| Reexamination appeal | Prosecution | Patent owner and examiner |
Ex parte appeals are the larger volume and the less discussed. Every applicant who receives a final rejection and does not amend or abandon can appeal, and those appeals go here.
Post-grant trials are the ones that reshaped patent litigation. They gave accused infringers a cheaper, faster route to challenge validity than a district court case.
Who decides
| PTAB | District court | |
|---|---|---|
| Decision-maker | Three administrative patent judges | Judge, often a jury |
| Technical background | Required | Not required |
| Legal qualification | Required | Required |
| Panel size | Three | One |
| Specialisation | Patents only | General docket |
Administrative patent judges must have both legal knowledge and scientific ability. In practice most hold technical degrees and many practised as patent attorneys before appointment.
That is a real difference for obviousness questions, which turn on what a person of ordinary skill in the art would have understood — a judgement a technically trained panel is better placed to make than a lay jury.
Volume and composition
| Proceeding type | Relative volume |
|---|---|
| Ex parte appeals | Largest |
| Inter partes review | Substantial |
| Post-grant review | Small |
| Derivation | Very small |
Ex parte appeals dominate by count and receive the least attention, because they are prosecution rather than litigation.
Post-grant review is small because of its nine-month window. Most disputes arise long after that closes. See post-grant review.
The standard of proof
| Forum | Standard for invalidity |
|---|---|
| PTAB post-grant trials | Preponderance of the evidence |
| District court | Clear and convincing evidence |
This single difference explains most PTAB strategy. An issued patent is presumed valid under §282, and overcoming that presumption in court requires clear and convincing evidence. Before the Board it does not.
The same prior art can therefore succeed at the PTAB and fail in court, which is why a defendant with strong documentary art goes to the Board where the grounds qualify.
Claim interpretation was aligned in 2018. The Board moved from broadest reasonable interpretation to the Phillips standard used in district court, which removed an inconsistency that had allowed the same claim to be read two ways.
How a trial runs
| Stage | Timing |
|---|---|
| Petition filed | Day 0 |
| Patent owner preliminary response | ~3 months |
| Institution decision | ~6 months |
| Patent owner response and motion to amend | ~3 months after institution |
| Petitioner reply | ~3 months later |
| Oral hearing | ~10–11 months after institution |
| Final written decision | Within 12 months of institution |
| Appeal to the Federal Circuit | Optional |
The statutory deadline is what makes the timetable reliable. A final written decision must issue within one year of institution, extendable by six months for good cause, and that certainty is part of the attraction.
Discovery is far narrower than in district court, which is where most of the cost saving comes from. There is no broad document production and no lengthy deposition programme.
Ex parte appeals
The larger half of the Board's work, and rarely discussed.
| Stage | Detail |
|---|---|
| Trigger | Final rejection by an examiner |
| Notice of appeal | Filed with a fee |
| Appeal brief | Argues the rejections are wrong |
| Examiner's answer | Responds |
| Reply brief | Optional |
| Oral hearing | Optional, on request |
| Decision | Affirm, reverse, or affirm in part |
Reversal returns the application to the examiner with the rejection overturned, which usually means allowance follows.
Appeal is slow. It adds a year or more to prosecution, which is why it is used where the examiner's legal position is wrong rather than where the prior art is simply strong. See patent application process.
Institution, and why it is the pivot
| Proceeding | Threshold |
|---|---|
| Inter partes review | Reasonable likelihood of prevailing on at least one claim |
| Post-grant review | More likely than not, or a novel legal question |
Institution is discretionary, not automatic. Meeting the threshold does not guarantee a trial, because the Board may decline on other grounds.
| Period | Approximate institution rate |
|---|---|
| October 2024 | ~65% |
| February 2026 | ~37% |
Always cite that figure with its date. The movement reflects discretionary practice rather than any change in petition quality, and it has been volatile enough that an undated number misleads.
A denial ends the matter without a decision on the merits, and the petition cost is spent. See PTAB discretionary denial.
Real party in interest
| Requirement | Why |
|---|---|
| Name every real party in interest | Statutory |
| Name privies | Statutory |
| Failure to name | Can defeat the petition |
| Purpose | Prevents evading the one-year bar through a proxy |
A party barred from filing cannot use another entity to file for them. Naming requirements exist to stop exactly that.
Getting it wrong is fatal rather than curable in some circumstances, which makes it a genuine diligence point rather than a formality.
Estoppel
| After a final written decision, the petitioner cannot raise | Where |
|---|---|
| Grounds raised | District court, ITC |
| Grounds that reasonably could have been raised | District court, ITC |
Estoppel attaches on a final written decision, not on institution or on filing. A petition that is denied institution carries no estoppel.
"Reasonably could have been raised" is read broadly, covering art the petitioner knew of or could have found with reasonable diligence. Filing therefore commits the documentary invalidity case.
Amending claims
| Position | |
|---|---|
| Motion to amend available | Yes |
| Historical success rate | Low |
| Pilot programme | Provides preliminary guidance on proposed amendments |
| Effect of amendment | Substitute claims, narrower |
| Intervening rights | May apply to amended claims |
The low success rate has been a persistent criticism, since the ability to amend was part of the original justification for the system — a patent owner facing invalidation should be able to narrow rather than lose everything.
The pilot giving preliminary guidance was introduced to address that, letting owners see the Board's initial view before committing.
The patent owner's toolkit
| Tool | When |
|---|---|
| Preliminary response | Before institution — the main opportunity |
| Discretionary denial arguments | Preliminary response |
| Motion to amend | After institution |
| Expert declaration | Post-institution response |
| Settlement | Any time, best before a final decision |
| Appeal | After a final written decision |
The preliminary response is where most owner wins happen, because a denial ends the matter without a merits decision and without estoppel for the petitioner.
Motions to amend need to be planned early. The historical success rate is low, and owners who treat it as a fallback rather than a strategy rarely succeed.
Stays of district court litigation
| Factor courts weigh | Favours a stay when |
|---|---|
| Stage of the case | Early |
| Simplification of issues | The IPR could resolve validity |
| Undue prejudice | The patent owner is not a competitor |
| Institution granted? | Yes — stays are far more likely after institution |
A stay changes the economics entirely. District court spending pauses while the Board proceeds on a fixed timetable, and the party bearing the discovery burden benefits most.
Which makes institution doubly important. It decides whether the challenge proceeds and, in practice, whether the parallel litigation pauses. See patent litigation.
Multiple petitions and joinder
| Situation | Treatment |
|---|---|
| Two petitions, same patent, same petitioner | Ranking and justification expected |
| A later petitioner joining an instituted trial | Joinder available |
| Joinder request timing | Within a month of institution in the trial joined |
| Effect on the one-year bar | Joinder can permit an otherwise time-barred party |
Joinder is a genuine route around the one-year bar in limited circumstances, allowing a party to join a trial already instituted on the same patent.
It is not a general workaround. The joining petition must be substantively similar, and discretion applies.
What the PTAB does not do
| Question | Decided by |
|---|---|
| Is the patent infringed? | District court or the ITC |
| What damages are owed? | District court |
| Should an injunction issue? | District court |
| Is the patent valid? | PTAB, or a court |
| Should this application be granted? | Examiner, then PTAB on appeal |
Validity and infringement are separate questions. A patent upheld at the Board may still not be infringed, and a patent that clearly reads on a product may still be invalid.
Worked example: choosing the forum
A company sued on one patent, holding two strong printed publications.
| District court only | Petition for IPR | |
|---|---|---|
| Invalidity standard | Clear and convincing | Preponderance |
| Decision-maker | Jury | Three technical judges |
| Cost of the validity fight | Part of a seven-figure case | Substantially less |
| Timeline | 2–4 years | ~18 months |
| Estoppel | None | Yes, on a final decision |
| Stay of the case | — | Likely if instituted |
| Institution risk | — | Real |
The decision
| Factor | Weight |
|---|---|
| Both references are printed publications | Grounds qualify |
| No §101 or §112 argument held back | Estoppel costs little |
| Discovery burden falls on the company | Stay is valuable |
| Institution rate uncertainty | Priced in |
| Outcome | Petition filed |
The estoppel was cheap here because there was nothing else to give up. A challenger holding a strong §101 argument would weigh it differently, since that ground was never available at the Board and is worth preserving for court.
The stay was worth as much as the merits. Pausing discovery for eighteen months shifts cost and leverage regardless of how the validity question resolves.
Appeals from the Board
| Issue | Standard on appeal |
|---|---|
| Factual findings | Substantial evidence |
| Claim construction from intrinsic evidence | De novo |
| Obviousness conclusion | De novo, on underlying facts reviewed for substantial evidence |
| Institution decisions | Generally not appealable |
Institution decisions are largely unreviewable, which concentrates a great deal of significance in a decision that cannot be challenged.
Claim construction is reviewed fresh, so a decision that turned on how a term was read is genuinely open at the Federal Circuit.
Practice materials
| Resource | Contains |
|---|---|
| Trial Practice Guide | Procedures, expectations, page limits |
| Precedential decisions | Binding on panels |
| Informative decisions | Persuasive, not binding |
| Standard operating procedures | How panels and reviews are managed |
Precedential designation matters. A decision designated precedential binds subsequent panels, which is how the Board develops consistent practice on discretionary questions.
The Trial Practice Guide is updated periodically, and checking the current version before filing is basic diligence.
Practical points for each side
| As petitioner | As patent owner |
|---|---|
| File early, not near the bar | Consider a preliminary response |
| One strong petition per patent | Argue discretionary denial |
| Name every real party in interest | Check the petitioner's naming |
| Use art the examiner never saw | Emphasise previously considered art |
| Seek a stay of the parallel case | Oppose the stay |
| Price the estoppel | Consider a motion to amend early |
The preliminary response is the owner's main pre-institution tool, and discretionary arguments are frequently where it does most work.
A motion to amend needs planning from the start, not as a late reaction, given how rarely they succeed without preparation.
PTAB: the checklist
- Distinguish the two halves. Ex parte appeals are prosecution; post-grant trials are adversarial.
- Check whether your grounds qualify. IPR covers only §102 and §103 on patents and printed publications.
- Use the preponderance standard where you can. It is the main advantage over district court.
- Diarise the one-year bar from service of an infringement complaint.
- Price the estoppel before filing, since it commits your documentary case.
- Treat institution as a real risk, not a formality, and note that rates move.
- Cite institution rates with their date. They are driven by discretionary practice.
- Consider the stay value separately from the merits. It can be worth as much.
- Remember the Board does not decide infringement, so a win there does not end the case.
- Plan for appeal. A final written decision is reviewable, and claim construction is reviewed de novo.