How Many Rejections Come Before a Grant?
One round. That is the median. Across 998,252 granted US utility applications filed 2020 to 2022, half saw exactly one rejection round before allowance and one in five saw none at all. About one in eight saw three or more.
A round here is one Office action containing a rejection, counting non-final and final actions. It is not one rejected claim, one statutory ground or one cited reference: a single action routinely carries several of each.
Round after round of rejections is the folklore. For applications that were eventually granted, it describes the tail rather than the middle.
The unit matters, so it is worth stating before the numbers. A round is one Office action that contains a rejection. One such action can reject several claims under several statutes over several references, and that still counts as one round here.
Granted applications only, which is the selection that matters here. Applications that were abandoned saw more: a median of two, and 15.0% saw four or more.
What the rejections are about
Prior art leads in every technology center. What differs between fields is what comes with it.
| Ground | All centers | Biotechnology |
|---|---|---|
| Prior art, section 102 or 103 | 87.1% | 78.2% |
| Obviousness, section 103 | 75.5% | 67.2% |
| Novelty, section 102 | 41.9% | 34.7% |
| Written description or definiteness, section 112 | 38.6% | 65.6% |
| Double patenting | 14.8% | 37.9% |
| Eligibility, section 101 | 14.5% | 13.2% |
| Prior art alone, with nothing else | 43.8% | 17.4% |
The last row is what separates the fields. In communications most prior-art rejections arrive on their own; in biotechnology they usually arrive with a section 112 rejection beside them, and only 17.4% of actions are prior art alone.
Which grounds come back in the next action
Not every ground gets argued round after round. Obviousness nearly always returns. The others come back about half the time.
The next action was itself a final in 85.9% of cases. Obviousness also arrives new, appearing in 34.2% of next actions where it was absent from the first; no other ground does that in more than 15%.
Field changes this sharply. Eligibility is cited again in 76.9% of cases in transportation and e-commerce against 22.0% in networks, and section 112 returns in 59.7% in biotechnology.
This compares ground against ground, not the same rejection over the same references, and the data does not say why a ground dropped away. Counted by matching the action text, calibrated against the USPTO structured labels and a reading of 300 actions.
Our own counts from the USPTO record. These describe what happened across past applications. They say nothing about any particular application, they are associations rather than causes, and none of this is legal advice.
The figures on this page are available as a CSV: /data/how-many-rejections-before-a-grant.csv. It is generated from the tables and charts above, so it carries the same numbers and cannot fall out of step with them.