Patent prosecution vs litigation is a distinction people get wrong in both directions, and the word prosecution is most of the problem.
Prosecution has nothing to do with criminal law. It is the administrative process of negotiating a patent application through a patent office.
Litigation is enforcement, in federal court, after a patent has issued.
The connection between them is one-way and consequential. Everything argued during prosecution becomes public and constrains what the claims can later be said to mean — so the cheapest stage decides the outcome of the most expensive one.
The core differences
| Prosecution | Litigation | |
|---|---|---|
| Forum | USPTO | Federal district court |
| Parties | Applicant and examiner | Patent holder and accused party |
| Nature | Administrative negotiation | Adversarial |
| Goal | Obtain a patent | Enforce or invalidate one |
| Claim interpretation | Broadest reasonable interpretation | Phillips — ordinary meaning |
| Invalidity standard | Preponderance | Clear and convincing |
| Decided by | Examiner | Judge or jury |
| Typical cost | $8,000–$20,000 | Seven figures |
| Typical duration | 2–4 years | 2–4 years, plus appeal |
| Who does it | Patent attorneys and agents | Litigators |
The two standards are the most consequential difference. During examination an examiner rejects on a preponderance; in court, an accused party must prove invalidity by clear and convincing evidence because the patent is presumed valid.
Which means the same prior art can behave differently depending on where it is raised — and is a large part of why inter partes review became so important.
Claim interpretation
| Standard | Where | Effect |
|---|---|---|
| Broadest reasonable interpretation | During examination | Makes rejection easier |
| Phillips | District court, and PTAB since 2018 | Ordinary meaning to a skilled person |
BRI during examination is deliberate. Reading claims broadly means the examiner tests them against more prior art, and the applicant can amend in response — which is not available after grant.
The PTAB moved to Phillips in 2018, aligning post-grant proceedings with district court and removing an inconsistency that had allowed the same claim to be read two ways.
The vocabulary problem
"Prosecution" is the word that causes most of the confusion, and it is worth being explicit about what it does not mean.
| Term | In patents | Elsewhere |
|---|---|---|
| Prosecution | Obtaining a patent from the USPTO | Criminal proceedings |
| Prosecutor | — | A government lawyer |
| Office action | An examiner's written rejection | Nothing related |
| Rejection | Applied to claims in prosecution | — |
| Invalidity | Applied to issued patents in litigation | — |
| Violation | Layperson term for infringement | — |
Claims are rejected; patents are held invalid. Using them interchangeably marks a document as coming from outside the field, and the distinction is real — rejection happens before grant, invalidity after.
There are no prosecutors in patent prosecution. The other party is an examiner applying statutory requirements, and they can be persuaded.
Who does the work
| Patent attorney | Patent agent | Litigator | |
|---|---|---|---|
| Technical degree | Required | Required | Not required |
| USPTO patent bar | Required | Required | Not required |
| Law licence | Yes | No | Yes |
| Prosecution | Yes | Yes | No |
| Litigation | Yes | No | Yes |
| Typical cost | Higher | Lower | Highest |
A general litigator cannot prosecute a patent application. Practising before the USPTO in patent matters requires registration, which requires a qualifying technical degree.
A patent agent cannot litigate, but does the same prosecution work as an attorney at lower cost. See patent attorneys.
Litigation teams frequently pair the two skill sets, because courtroom experience and deep technical prosecution knowledge rarely sit in one person.
Two different documents
Prosecution produces a negotiated document; litigation produces a judgment about it.
| Prosecution output | Litigation output | |
|---|---|---|
| Form | An issued patent | A judgment or settlement |
| Negotiated with | An examiner | The other party |
| Public record created | The file wrapper | The docket |
| Binding on | The applicant, going forward | The parties |
| Reviewable | By the PTAB, then the Federal Circuit | By the Federal Circuit |
The file wrapper is the durable artifact. Everything an applicant said to get a patent stays public and is read by every party who ever considers the patent afterwards — buyers, licensees and accused infringers alike.
Which makes prosecution a form of drafting for an audience that has not appeared yet. The examiner is the immediate reader; the eventual reader is opposing counsel.
How prosecution constrains litigation
This is the connection that matters, and it runs one way.
| Prosecution act | Litigation consequence |
|---|---|
| Amending a claim to overcome prior art | Prosecution history estoppel — that scope is generally surrendered |
| Arguing a term means something specific | Binds the later construction |
| Characterising prior art in the background | Can be an admission |
| Accepting an examiner's characterisation | Binds you |
| Inconsistent terminology | §112 indefiniteness exposure |
| Narrow claims accepted to secure allowance | Little left to enforce |
| Few dependent claims | No fallback when the independent falls |
Every amendment is public and permanent. The file wrapper is the first thing an accused party's counsel reads, because it constrains what the patent holder can now argue the claims mean.
Which is the real argument for careful prosecution. A claim narrowed unnecessarily to end an argument quickly costs nothing that year and can cost the case a decade later.
Where the PTAB sits
Across both, which is why it confuses the categories.
| PTAB function | Closer to |
|---|---|
| Ex parte appeal from an examiner rejection | Prosecution |
| Inter partes review | Litigation |
| Post-grant review | Litigation |
| Derivation proceedings | Prosecution |
Appeals from examiners are the tail end of prosecution. The applicant is still trying to obtain a patent, just before a different decision-maker.
IPR is adversarial and post-grant. Two parties, a challenger and a patent owner, and the outcome cancels claims. See inter partes review.
The preponderance standard applies to both, which is what makes the PTAB attractive to challengers relative to district court.
Where each one can go wrong
| Prosecution failure | Consequence |
|---|---|
| Claims too narrow | Nothing worth enforcing |
| Unnecessary amendment | Estoppel on surrendered scope |
| Few dependent claims | No fallback when the independent falls |
| Specification lacks alternatives | Cannot amend — new matter barred |
| Inconsistent terminology | §112 exposure |
| Missed disclosure duty | Unenforceability |
| Litigation failure | Consequence |
|---|---|
| Adverse claim construction | Case effectively over |
| Weak prior art in contentions | Invalidity defence fails |
| Late references excluded | Best art unusable |
| No marking compliance | Pre-notice damages lost |
| Missed one-year IPR bar | Route forfeited |
| Willfulness finding | Damages trebled |
Prosecution failures are permanent; litigation failures are case-specific. A badly drafted specification cannot be fixed, because new matter cannot be added after filing.
Cost and duration compared
| Stage | Prosecution | Litigation |
|---|---|---|
| Entry cost | Search and drafting, $6,500–$18,000 | Pre-suit investigation, tens of thousands |
| Middle | Office action responses, $2,000–$4,000 each | Discovery, hundreds of thousands |
| Decision point | Notice of allowance | Claim construction |
| Total | $8,000–$20,000 | $1m+ through trial |
| Duration | 2–4 years | 2–4 years, plus appeal |
The multiple is roughly fifty to one, and the whole of the difference is adversarial process — discovery, experts, motions.
Both have a pivot point. In prosecution it is the first office action response, where claim scope is largely settled. In litigation it is claim construction, where the outcome is largely settled. See patent litigation.
Different work, different rhythm
| Prosecution | Litigation | |
|---|---|---|
| Pace | Slow, months between actions | Intense, deadline-driven |
| Adversary | An examiner applying rules | An opponent seeking to win |
| Written work | Amendments, arguments | Briefs, motions, reports |
| Oral advocacy | Examiner interviews | Hearings, depositions, trial |
| Technical depth | Very high | High, mediated through experts |
| Outcome | A negotiated document | A judgment |
An examiner is not an opponent. They apply statutory requirements and can be persuaded by a well-reasoned argument or a twenty-minute interview, which is a very different dynamic from opposing counsel.
Examiner interviews are the most underused tool in prosecution, and the closest thing it has to oral advocacy. See patent prosecution.
Skills that do not transfer
| Prosecution requires | Litigation requires |
|---|---|
| Deep technical fluency | Technical fluency, mediated through experts |
| Claim drafting | Claim construction argument |
| Knowledge of examiner practice | Knowledge of local rules and judges |
| Patience across years | Intensity across deadlines |
| Persuading a technical reader | Persuading a generalist judge or jury |
| USPTO registration | Court admission |
Persuading an examiner and persuading a jury are different crafts. An examiner is a technical specialist who will read a detailed argument; a jury is not.
This is why teams pair the two. A prosecution specialist who understands why the claim reads a certain way, working with a litigator who can present it, is the common arrangement.
What each side reads first
| Role | Reads first |
|---|---|
| Examiner | The claims, then searches prior art |
| Accused party's counsel | The file wrapper |
| Patent holder's counsel | The accused product, then the claims |
| PTAB panel | The petition and the prior art |
| Judge at Markman | The intrinsic record — claims, specification, file history |
The file wrapper being read first by the other side is the practical consequence of everything above. Prosecution creates the record that enforcement is argued from.
Worked example: one decision, two effects
During prosecution, an examiner rejects claim 1 over two references.
| Option | Prosecution effect | Litigation effect |
|---|---|---|
| Argue without amending | Slower, may fail | No estoppel — full scope preserved |
| Amend narrowly | Faster allowance | Estoppel on the surrendered scope |
| Amend broadly to end it quickly | Fastest | Large surrender; little left to enforce |
| Appeal | Slowest, costly | No surrender if successful |
Ten years later
| Argued, no amendment | Amended narrowly | |
|---|---|---|
| Claim scope | Broader | Narrower |
| Competitor's product | Within the claim | Outside it |
| Equivalents available | Yes | Estopped |
| Outcome | Enforceable | Nothing to assert |
The decision cost a few thousand dollars and six months in year two. It decided whether the patent was worth anything in year twelve.
Which is why prosecution deserves litigation-grade thinking, even though it feels administrative at the time.
Moving between the two
Some decisions belong to both, and they are worth identifying early.
| Decision | Made during | Felt during |
|---|---|---|
| Claim scope | Prosecution | Litigation |
| Continuation strategy | Prosecution | Litigation |
| Marking compliance | Product launch | Damages |
| Entity status | Fee payment | Neither |
| Documenting secondary considerations | Prosecution | Both |
Continuations are the clearest example. A continuation kept pending lets claims be written toward a competitor's product once it exists, which is a litigation advantage created entirely during prosecution.
Most patents never reach litigation
| Ipiry Patent Survival Curve v1.0 | Rate |
|---|---|
| Survive the 3.5-year fee (2022 cohort) | 85.8% |
| Survive the 7.5-year fee (2018 cohort) | 64.6% |
| Reach full term (2014 cohort) | 41.4% |
| Abandoned before full term | 58.6% |
Computed from 27,273,654 USPTO maintenance fee records covering 8,262,336 US utility patents — see the patent survival curve.
Nearly three in five are abandoned before term, and only a small fraction of the remainder are ever asserted. Prosecution is the stage almost every patent goes through; litigation is the stage almost none reach.
Which does not make prosecution less important. It makes it the only stage where the decisions are certain to matter.
Reissue and correction
A third category sits between the two — fixing an issued patent.
| Mechanism | Purpose | Timing |
|---|---|---|
| Certificate of correction | Clerical or typographical errors | Any time |
| Reissue | Correcting a defective patent | Broadening within 2 years |
| Disclaimer | Surrendering claims | Any time |
| Supplemental examination | Considering information not previously before the office | Any time |
Broadening reissue has a two-year window from the original grant, and it is one of the few chances to widen claims after issue.
Reissue is prosecution work on an issued patent, which is why it sits between the two disciplines and is usually handled by prosecution counsel.
Prosecution vs litigation: the checklist
- Do not confuse patent prosecution with criminal prosecution. They share only a word.
- Understand the two claim interpretation standards — BRI during examination, Phillips in court and at the PTAB.
- Note the invalidity standards differ — preponderance at the PTAB, clear and convincing in district court.
- Use a patent agent for prosecution where litigation is not anticipated; the work is identical and costs less.
- Treat every prosecution amendment as a litigation decision, because estoppel is permanent.
- Argue before amending where the argument is available, to preserve scope.
- Include generous dependent claims. They are fallbacks in both prosecution and litigation.
- Watch the background section for admissions that will be read back to you.
- Read the file wrapper before asserting, because the other side certainly will.
- Budget realistically. Litigation costs roughly fifty times what obtaining the patent did, which is why prosecution quality is the cheapest leverage available.