CPA Exam
pass rates
The CPA Exam pass rate by section for 2025, cumulative: 48.21% for AUD, 42.12% for FAR, 63.12% for REG, 41.94% for BAR, 67.79% for ISC and 77.65% for TCP. FAR and BAR are the two hardest sections to pass by this measure, and FAR is the lowest of the three Core sections. Every figure on this page was read off the AICPA’s own pass-rate page on 2026-09-05; where we could not reach a primary source, the row says so rather than dropping the flag.
Source: UWorld Accounting, “CPA Exam Pass Rates”, compiling AICPA quarterly reports · checked 2026-08-15
CPA Exam pass rate by section, 2025 and 2026
| Section | 2025 cumulative | 2026 Q1 | 2026 Q2 | 2026 cumulative |
|---|---|---|---|---|
| AUD | 48.21% | 47.80% | 49.38% | 48.65% |
| FAR | 42.12% | 43.46% | 42.47% | 42.95% |
| REG | 63.12% | 66.65% | 66.87% | 66.78% |
| BAR | 41.94% | 41.30% | 45.62% | 43.64% |
| ISC | 67.79% | 66.79% | 67.98% | 67.45% |
| TCP | 77.65% | 79.28% | 79.75% | 79.53% |
Two readings worth separating. A low pass rate is not a difficulty ranking. Sections differ in who sits them and when in a candidate’s sequence they are taken, so the spread between 42% and 78% is not a statement that one body of material is 36 points harder than another. And a cumulative figure is not your probability. It is the base rate across every attempt, which is where an estimate should start and not where it should end.
The same table answers the question people usually ask instead — which section is hardest. By 2025 cumulative pass rate the order is BAR, FAR, AUD, REG, ISC, TCP, hardest first. Read it as where attempts fail, not as where the material is hardest: the two are different claims and only the first one is in the data.
Machine-readable twin of this table, section by section and quarter by quarter, with the same flags: /data/pass-rates.json. Quote it with attribution to passfar.com and please carry the flag with the number.
Move that base rate with your own evidence → · What FAR actually tests, group by group →
FAR by quarter
| Year | Q1 | Q2 | Q3 | Q4 | Cumulative |
|---|---|---|---|---|---|
| 2024 | 41.92% | 40.58% unverified | 39.82% unverified | 36.80% unverified | 39.59% unverified |
| 2025 | 41.67% | 43.52% | 43.07% | 40.20% | 42.12% |
| 2026 | 43.46% | 42.47% | — | — | 42.95% |
Read the flags. Only the 2025 cumulative figure is corroborated across several independent compilations. Every other row comes from a single third-party compilation of AICPA quarterly reports and has not been checked against a primary AICPA or NASBA release. We publish them because withholding data is its own kind of dishonesty — but do not quote an unverified row without the flag.
Machine-readable, with the same flags: /data/pass-rates.json
CPA Exam retake pass rate: do retakers pass less than first-time candidates?
Almost certainly yes — and we are not going to put a number on it, because there is no free primary source that publishes one. The CPA Exam retake pass rate is not in the AICPA’s quarterly tables, which report one rate per section per quarter across all candidates. The first-time-versus-repeat split lives in NASBA’s annual Candidate Performance on the Uniform CPA Examination, a priced publication we have not bought.
Figures in the 10–15 point range circulate widely and are attributed to NASBA. We checked on 2026-09-05 and could not find one on a public NASBA page, so we do not reprint it. A statistic you cannot follow to its source is one you should not quote, including from us.
What a gap would not establish is why. The obvious reading — that failing is an execution problem rather than a knowledge problem, since a repeat candidate has already studied the material once — is a real hypothesis, and it is the one this product was built around. It is also not what such a statistic would show. The pool of repeat candidates is selected: it is, by construction, the candidates who did not pass the first time. A group assembled that way would be expected to perform below average on the next attempt even if nothing about how they were studying mattered at all.
So the honest statement is the narrow one: a re-sit is not the easier road. We use the wider claim internally as a hypothesis to design against. We do not print it as a finding, and you should be suspicious of anyone who does.
We flag this because the retaker gap is the most quotable number in CPA prep, the most frequently over-read one, and the one most often quoted without a source.
Common questions
What is the CPA Exam pass rate?
It differs by section. For 2025 the cumulative rates were AUD 48.21%, FAR 42.12%, REG 63.12%, BAR 41.94%, ISC 67.79% and TCP 77.65%. There is no single CPA Exam pass rate, and quoting one usually means quoting whichever section makes the point.
Which CPA Exam section has the lowest pass rate?
On 2025 cumulative figures, BAR at 41.94% and FAR at 42.12% are the two lowest, within a fifth of a point of each other. Among the three Core sections FAR is the lowest.
What is the FAR pass rate?
42.12% cumulative in 2025, and 42.95% across the first two quarters of 2026. Roughly six in ten attempts do not pass.
Which CPA Exam section is the hardest?
Ranked by 2025 cumulative pass rate the order is BAR 41.94%, FAR 42.12%, AUD 48.21%, REG 63.12%, ISC 67.79%, TCP 77.65% — so BAR and FAR are where the most attempts fail. That is the honest answer to the question people are asking, and it is not the same as a difficulty ranking; see the next answer.
Does a low pass rate mean the section is harder?
Not on its own. Sections differ in who sits them and where they fall in a candidate’s sequence, so the gap between sections mixes difficulty with who is taking the exam. It is a base rate, not a difficulty ranking.
Do retakers pass less than first-time candidates?
Almost certainly yes, but we will not put a number on it. The AICPA’s published tables give one rate per section per quarter across all candidates; the CPA Exam retake pass rate — the first-time-versus-repeat split — is in NASBA’s priced annual candidate performance report, which we have not bought, and we could not find it on a public page when we checked on 2026-09-05. Note also that the comparison is between two groups that were not assembled the same way: the repeat pool is by construction the candidates who did not pass the first time, and a group selected that way would be expected to sit below average again even if nothing about their studying mattered. A gap would establish that retaking is harder. It would not be a diagnosis of why.
Are these numbers official?
The 2025 and 2026 figures on this page were read off the AICPA’s own pass-rate page on 2026-09-05. The 2024 quarterly rows come from a secondary compilation and stay flagged, because the AICPA has taken those quarterly pages down and we cannot reach a primary source for them.
What we would need to say more
A pass rate is a population statistic. Turning one into a statement about a specific candidate needs a link between practice performance and exam outcomes, and that link is exactly what nobody in this market publishes. We do not have it either. The estimator on this site starts from the population rate and moves away from it in proportion to evidence — which is the correct thing to do when you have no calibration data, and not a substitute for having it.