Inter partes review: what an IPR is and when to file one

Inter partes review is an adversarial proceeding before the Patent Trial and Appeal Board that cancels claims of an issued patent as anticipated or obvious over prior art. It was created by the America Invents Act as a faster, cheaper alternative to litigating validity in district court.

One note on terminology before anything else. IPR in US patent practice means inter partes review. It also commonly means intellectual property rights generally, and in Indian practice it means something different again. In writing aimed at a mixed audience, spell it out.

What an IPR can and cannot do

The limitation to prior art patents and printed publications is the defining constraint, under 35 U.S.C. 311(b). A devastating public-use case cannot be raised at the Board at all, which means the prior art search that supports an IPR petition has to be documentary from the outset.

The burden is lower than in court. The Board applies a preponderance of the evidence standard, while a district court requires clear and convincing evidence to invalidate. That difference alone makes the Board a more favourable forum for a challenger with reasonable art.

The deadlines that decide whether you can file at all

The one-year bar is the one that catches people. It runs from service of a complaint alleging infringement, and it applies not only to the named defendant but to real parties in interest and privies. A supplier indemnifying a customer who was served fourteen months ago may find the bar already run.

Diarise it on the day of service. By the time a defendant has assessed the case, engaged counsel and commissioned a prior art search, several months of the year have gone. A petition needs the art, the expert declaration and the claim construction positions ready before filing.

The nine-month wait applies only to AIA patents — those with an effective filing date on or after 16 March 2013. For those, post-grant review occupies the first nine months and IPR becomes available afterwards.

What it costs

USPTO fees effective January 2025.

Official fees are the smaller half. Attorney fees, expert declarations and the prior art search typically exceed them, and the total for a contested IPR through final written decision runs well into six figures.

It remains materially cheaper than district court invalidity litigation, which is the entire economic rationale. A defendant facing a claim worth less than the cost of defending it in court has a proceeding available that costs substantially less.

Note the fee structure incentive. Challenging more than twenty claims raises the cost sharply at both stages, which encourages petitioners to target the claims actually asserted rather than the whole patent.

The timeline

Statutory deadline for the final written decision is one year from institution, under 35 U.S.C. 316(a)(11), extendable by six months for good cause. In practice the Board meets it.

Total elapsed time from petition to decision is commonly around eighteen months — roughly six to institution and twelve after. That is fast by litigation standards and is why district courts frequently stay parallel cases pending the outcome.

Institution: the decision that matters most

Institution is discretionary and the odds have moved sharply.

That is a fall of more than 40% in relative terms, driven principally by changes to discretionary denial practice. Monthly figures within that period have been volatile, so any institution statistic is only meaningful with its date attached — a rate quoted from 2023 tells you very little about filing today.

Two kinds of denial exist and they are different animals.

Merits denial means the petition failed to show a reasonable likelihood that at least one challenged claim is unpatentable. That is a judgement about the prior art.

Discretionary denial means the Board declined to institute despite the merits, most often because parallel district court litigation is likely to reach validity first. That is a judgement about efficiency, and it has nothing to do with the strength of the art.

The practical consequence for petitioners. A strong petition can now be denied for reasons unrelated to its quality, which changes the expected value of filing. It also increases the importance of filing early — a petition filed promptly after service faces a weaker argument that the district court will get there first.

Estoppel: the price of trying

This is the part that is easy to underweight and expensive to discover late.

Under 35 U.S.C. 315(e), once the Board issues a final written decision, the petitioner is estopped from asserting in district court or at the ITC any ground that was raised or reasonably could have been raised during the IPR.

"Reasonably could have been raised" is broad. It covers prior art the petitioner did not use but could have found with a diligent search — not only the references actually cited.

The asymmetry is stark. A denied petition costs money and forecloses nothing. An instituted petition that fails costs money and removes your best invalidity defences from the parallel litigation.

What estoppel does not reach. Public use, on-sale activity, section 101 and section 112 grounds were never available in the IPR, so they survive. That is worth knowing before filing, because a defendant whose strongest case is a public use may be better off in court entirely.

A worked decision: IPR or district court

A defendant served with a complaint in March. The asserted patent has fifteen claims, eight of them asserted. The prior art search produces one strong documentary reference the examiner never saw, plus evidence that a competitor publicly demonstrated the invention two years before the priority date.

The public demonstration changes the calculation. It is likely the strongest invalidity case and it cannot be raised at the Board. Filing an IPR on the documentary art alone risks losing, triggering estoppel on all documentary grounds, and then litigating the public use in court without them.

A reasonable path here is to preserve the public use for court, assess whether the documentary reference alone justifies an IPR, and decide before the one-year bar rather than at it.

Change one fact and the answer flips. Remove the public demonstration and the documentary reference becomes the whole case — at which point the Board's lower burden and shorter timeline make it clearly the better forum, subject to institution risk.

Who files IPRs

Defendants in infringement litigation, most commonly, using the proceeding to resolve validity faster and cheaper than the parallel case.

Companies not yet sued. There is no standing requirement — any person other than the patent owner may petition. A competitor concerned about a patent can challenge it before any assertion occurs, and the one-year bar has not started running.

Defensive aggregators and industry groups, funding petitions against patents that threaten their members. Because cancellation is effective against the world rather than just the petitioner, a single successful IPR benefits everyone exposed.

This last point matters commercially. A patent surviving an IPR is worth more than one that has never been tested, which feeds directly into any patent valuation, because the challenge has been made and failed. A patent with claims cancelled is worth correspondingly less, and one under a pending petition is difficult to value or sell at all.

Before filing a petition

1. Diarise the one-year bar from the date of service, and work backwards.
2. Confirm no privy or real party in interest was served earlier, because the bar travels.
3. Establish that your best art is documentary — public use and on-sale evidence cannot be raised.
4. Search for what the examiner missed, since art already of record faces the argument it was considered.
5. Model the estoppel exposure if the petition is instituted and fails.
6. Check the current institution rate, not a figure from two years ago.
7. Assess whether a stay of the district court case is realistic, because that is often the actual objective.