A U.S. utility patent generally expires 20 years from its earliest non-provisional filing date, a term set by 35 U.S.C. § 154(a)(2). Because the clock runs from filing rather than from issuance, time consumed by the U.S. Patent and Trademark Office during examination eats directly into the period during which the patent can be enforced. Patent term adjustment, codified at § 154(b), is the statutory mechanism that returns some of that lost time when the delay is the Office's responsibility.

The statute frames the adjustment around three categories of USPTO delay, commonly labeled A, B, and C. The A guarantee, in § 154(b)(1)(A), is a "guarantee of prompt Patent and Trademark Office responses" — it sets specific deadlines such as providing a first office action or notice of allowance within 14 months of filing and responding to applicant submissions within 4 months. When the Office misses one of these, the statute provides that "the term of the patent shall be extended 1 day for each day" until the required action is taken.

"the term of the patent shall be extended 1 day for each day after the end of the period specified in clause (i), (ii), (iii), or (iv), as the case may be, until the action described in such clause is taken."— 35 U.S.C. § 154(b)(1)(A), source

The three-year guarantee and overlap

The B guarantee, in § 154(b)(1)(B), is a "guarantee of no more than 3-year application pendency." If the USPTO does not issue a patent within three years of the actual filing date — excluding time consumed by certain applicant-requested or statutory delays — the term is extended day-for-day for the excess. The C guarantee, in § 154(b)(1)(C), addresses delays from interferences, derivation proceedings, secrecy orders, and successful appeals. The Federal Circuit has shaped how these provisions operate — most consequentially in Novartis AG v. Lee and Supernus Pharmaceuticals, Inc. v. Iancu, which addressed how B-delay and applicant-delay offsets interact.

A key wrinkle is overlap. Section 154(b)(2)(A) provides that to the extent periods of delay under the A, B, and C guarantees overlap, the adjustment cannot exceed the actual total number of days the issuance was delayed. The same day of delay cannot be counted twice. This anti-double-counting rule was the subject of the Federal Circuit's Wyeth v. Kappos decision, which corrected the USPTO's earlier method of computing overlapping A and B delay.

Applicant delay reduces the award

PTA is not a one-way benefit. Section 154(b)(2)(C) requires the Office to reduce any adjustment by the period during which the applicant "failed to engage in reasonable efforts to conclude prosecution." The implementing regulations at 37 CFR 1.704 enumerate specific applicant actions — such as late responses, certain preliminary amendments, and supplemental papers — that count as applicant delay and are subtracted from the gross adjustment. An applicant who takes the full statutory response period or files late submissions can see its PTA shrink accordingly.

Procedurally, the USPTO calculates an initial PTA and prints it on the issued patent. An applicant who disagrees may request reconsideration within the statutory window and, under § 154(b)(4), may seek review in the U.S. District Court for the Eastern District of Virginia. PTA is distinct from patent term extension under § 156, which compensates for time lost to regulatory review by agencies such as the FDA and is governed by a separate statute and calculation. For most patents, however, term adjustment under § 154(b) is the operative mechanism: it restores effective term lost to examination delay, day for day, net of any delay the applicant itself caused.

Why effective term, not nominal term, is the stake

The economic reason PTA matters is that the 20-year clock under § 154(a)(2) runs from filing, so every month the USPTO spends examining an application is a month subtracted from the period of enforceable exclusivity. For technologies with long development timelines, that erosion can be material, and PTA restores the days lost to Office delay so that the patentee retains close to the full term Congress intended. The adjustment is printed on the face of the issued patent and becomes part of the patent's term of record.

The Federal Circuit's case law has repeatedly tightened how the USPTO must calculate the award. In Wyeth v. Kappos, the court rejected the Office's method of treating overlapping A-delay and B-delay, holding that overlap occurs only when the delays fall on the same calendar days. In Supernus Pharmaceuticals, Inc. v. Iancu, the court held that the Office could not deduct as applicant delay a period during which there was no action the applicant could have taken to advance prosecution. Each decision narrowed the circumstances in which PTA can be reduced, and together they illustrate that the calculation is rule-bound rather than discretionary.

For applicants, the practical lesson is procedural discipline. Because 37 CFR 1.704 converts a wide range of applicant conduct — late responses, certain supplemental replies, and specified filings — into deductions from the gross adjustment, the way an applicant prosecutes an application directly affects the term it ultimately receives. PTA rewards prompt prosecution and penalizes delay on the applicant's side, even as it compensates for delay on the Office's side, and the day-for-day arithmetic of § 154(b) makes the consequences of each side's timing precisely measurable.