A USPTO examiner approves a patent application only if the invention it claims is useful, novel, and non-obvious over existing patents and publications, known as "prior art." Patent examination is a search-and-compare exercise: the examiner hunts for the closest prior art, lines it up element by element against your claims, and issues either a notice of allowance or an office action naming each rejected claim and the reason.
Nobody builds your invention. Nobody tests it. The examiner reads what you wrote and asks whether the words of your claims cover something the public already had.
What happens to your application after you file
The first stage is clerical, not substantive. The USPTO checks that the application has the required parts, fees, and formal drawings, then issues a filing receipt and assigns a classification. That classification determines which technology center and art unit your case lands in, which in turn determines who examines it and how long you wait.
From there, your application joins a queue ordered largely by filing date within the art unit. The USPTO’s own process overview lays out the sequence from filing through examiner assignment to grant, but it does not convey how much of the calendar is simply waiting. Most of the elapsed time before your first office action is queue time, not examination time. Separately, utility applications publish about 18 months after the earliest filing date, whether or not an examiner has touched them.
When the case finally reaches the top of the docket, the examiner reads the claims first, then the specification and drawings to understand what the claim terms mean. Only then does the prior art search begin, across US and foreign patent databases, published applications, and non-patent literature. If the application appears to cover two or more distinct inventions, the examiner may issue a restriction requirement and make you pick one before substantive examination proceeds.
This staged structure, where a formal completeness check precedes a substantive scientific review, mirrors other federal approval pipelines such as how the FDA approves a new drug. The difference is that the patent examiner evaluates a document, not a product.
The three legal tests an examiner applies
Three statutory sections do nearly all the work. Section 101 of Title 35 covers patent eligibility and utility: the claimed thing has to fall into one of the statutory categories (process, machine, manufacture, or composition of matter) and have a specific, credible, real-world use. Abstract ideas, laws of nature, and natural phenomena are excluded, which is why 101 rejections cluster in software, business methods, and diagnostics.
Section 102 is novelty. A claim fails if a single prior art reference discloses every element of that claim, arranged as claimed. One reference, all elements. If the examiner has to combine two documents to cover your claim, that is not a 102 problem.
Section 103 is the harder one. Under 35 U.S.C. 103, a patent cannot be obtained if the differences between the claimed invention and the prior art would have been obvious, before the effective filing date, to a person having ordinary skill in that art. In practice the examiner cites two or more references and explains why someone skilled in the field would have combined them and had a reasonable expectation that the combination would work.
The practical consequence is worth spelling out. Novelty is a bright line you can test by reading one document. Obviousness is a judgment call about what a hypothetical skilled person would have found predictable, which is why most back-and-forth between applicants and examiners is a 103 argument. Beating a 103 rejection usually means attacking the reason for combining the references, or amending the claim to add a limitation that no cited reference teaches.
Two other requirements sit in Section 112: the specification has to describe the invention well enough to enable someone skilled in the field to make and use it, and the claims have to be definite. These rejections are about drafting quality rather than whether the invention is new.
How long examination actually takes right now
The authoritative numbers live on the USPTO Patents Dashboard, which defines first office action pendency as the average number of months from filing date to the mailing of the first office action, averaged over the most recent quarter. Traditional total pendency measures filing to final disposition, meaning either issuance or abandonment.
Published summaries of that data disagree at the margins, and it is worth knowing why. One compilation of USPTO patent filing statistics puts FY 2025 first action pendency at 20.5 months and FY 2024 total pendency at 26.1 months, while a 2026 timeline summary citing the May 2026 dashboard reports roughly 21.5 months to first action and 29 to 34 months of total pendency. The spread comes from fiscal-year versus quarterly snapshots, and from whether applications with a request for continued examination are folded in. The dashboard figure updates quarterly, so treat any number you read in a blog post as a rough anchor and check the dashboard for the current quarter.
Art unit matters more than the national average. Mechanical and manufacturing art units have historically run slower than communications art units, and a backlog of unexamined applications sitting in the high hundreds of thousands means the queue, not examiner speed, sets your wait.
If the timeline is the problem, Track One prioritized examination is the main lever. The USPTO fee schedule lists the prioritized examination fee under 37 CFR 1.17(c) at $4,515 for large entities, $1,806 for small entities, and $903 for micro entities, on top of normal filing, search, and examination fees. The USPTO’s goal is a final disposition within about 12 months, and in July 2025 the agency raised the annual cap on accepted Track One requests from 15,000 to 20,000. Final disposition means a decision, not an approval; a final rejection counts.
Office actions: what a rejection actually means
An office action is the examiner’s written reasoning plus a deadline. It identifies each claim, states the statutory basis for every rejection, names the prior art relied on, and tells you when your reply is due. It is not a denial of your application and it does not close the file.
The maximum statutory period for replying to an office action is six months under 35 U.S.C. 133, and as MPEP 710 notes, the USPTO sets a shorter period in practically every case. The usual shortened period is three months from the mail date, extendable one month at a time, for escalating fees, up to the six-month ceiling. Miss the last day and the application goes abandoned automatically.
Non-final rejection
The first substantive office action is normally non-final. You get the widest set of moves here: amend the claims, argue that the cited references do not disclose what the examiner says they do, submit evidence such as a declaration showing unexpected results, or any combination. The examiner must consider the amendment on the merits and issue a new action.
Final rejection
"Final" refers to this round of prosecution, not to the invention. After a final rejection, amendments are entered only at the examiner’s discretion, typically when they place the case in condition for allowance or merely cancel claims. Your real options narrow to a request for continued examination, an appeal to the Patent Trial and Appeal Board, or letting the case go. One compilation of USPTO data reports that more than 92% of applications receive at least one rejection before disposition, so a first-action rejection says very little about your eventual odds.
What happens next at each outcome
| Outcome | What it means | Your options | Clock |
|---|---|---|---|
| Notice of allowance | Examiner found claims patentable as they stand | Pay the issue fee; file a continuation first if you want broader or different claims | Issue fee due 3 months from the notice mail date, and that window cannot be extended (37 CFR 1.311) |
| Non-final rejection | Examiner’s first reasoned position; prosecution open | Amend claims, argue the art, add evidence, request an examiner interview | Usually 3 months, extendable to 6 for fees |
| Final rejection | This round is closed; amendments entered only at examiner’s discretion | File an RCE ($1,500 large, $600 small, $300 micro for the first request), appeal to the PTAB, file a continuation, or abandon | Usually 3 months, extendable to 6 for fees; a notice of appeal or RCE must land inside that window |
| Abandonment | No timely reply was filed; the application is dead | Petition to revive on the basis of unintentional delay under 37 CFR 1.137 | No fixed cutoff, but the fee is higher and more scrutiny applies after one year |
Two entries in that table deserve expansion. The RCE amounts come from the USPTO fee schedule under 37 CFR 1.17(e)(1); a second and subsequent RCE costs $2,860 for a large entity. An RCE reopens prosecution with the same examiner, which is the right move when you have a concrete claim amendment to try and the wrong move when the dispute is purely about whether the examiner’s legal reasoning holds.
Revival is not automatic. Under 37 CFR 1.137, a grantable petition must include the reply that was missed, the petition fee set in 37 CFR 1.17(m), and a statement that the entire delay was unintentional. The USPTO confirmed in a June 2026 Federal Register notice that it may require additional information where there is a question whether the delay was unintentional, with particular attention to petitions filed more than a year after abandonment.
Common misconceptions about patent examination
"Rejected means denied." It means the examiner has stated a position you can answer. Given that the overwhelming majority of applications draw at least one rejection, a first office action full of 103 rejections is the normal midpoint of the process, not the end of it.
"Patent pending protects me." A pending application gives you a priority date and, after publication, potential provisional rights, but you cannot sue anyone for infringement until a patent actually issues with claims that cover their product.
"The examiner will evaluate whether my invention works better." Commercial merit is not a patentability criterion, and no prototype is built or tested. The examiner evaluates claim language against documents. This is a narrower, more mechanical review than many applicants expect, in the same way that how Social Security disability claims get decided turns on documented criteria rather than an open-ended judgment about hardship.
"I need an attorney to file." You do not; individual inventors can prosecute their own applications. The reason most applicants hire a registered patent attorney or agent anyway is that claim scope, not invention quality, decides most cases, and claim drafting is a specialized skill. For anything with real commercial stakes, discuss your situation with a registered practitioner before responding to an office action.
Frequently asked questions
Can I talk to my examiner? Yes. Under 37 CFR 1.133 and MPEP 713, interviews are arranged through the Automated Interview Request form or form PTOL-413A, and an interview that would expedite allowance should be granted after a first action reply. Before the first office action, interviews are limited to continuing or substitute applications or cases where the examiner decides one would advance prosecution. The USPTO’s interview practice FAQs state that normally one interview is permitted after a final rejection.
Does money speed things up? Track One does, measurably, and nothing else reliably does for a standard utility filing. Petitions to make special on other grounds exist but carry their own requirements.
What if I miss the deadline? The application is abandoned by operation of law. Revival under 37 CFR 1.137 requires the missed reply, the petition fee, and the unintentional-delay statement.
Can I appeal a final rejection myself? Yes. PTAB appeals follow strict briefing rules and run on their own multi-month queue, which is why most applicants retain counsel at that stage.
A workable decision rule after a final rejection: if the examiner’s position rests on a claim limitation you are willing to narrow, an RCE is usually the faster path; if you believe the examiner has misread the cited reference or misapplied the obviousness standard and you are unwilling to narrow, an appeal puts that question in front of a different decision-maker. Price both against the remaining commercial life of the invention before choosing.