On this page
CFA Institute’s latest released CFA® Program results as of 25 October 2026 are for the August 2026 exam window at Levels I, II and III. It reports pass rates of 43%, 45% and 47% respectively. These rates describe the candidates who sat each level in that window; they do not reveal an individual candidate’s score, the minimum passing score, or the chance that a particular future candidate will pass.
Read the August window level by level
The three results came out on different dates, so “latest” needs a window and a release date. CFA Institute’s Level I announcement was dated 24 September 2026 and reported that 28,145 candidates sat the August Level I exam, with 43% passing. Its Level II announcement, dated 30 September 2026, reported 10,004 candidates and a 45% pass rate. The Level III announcement, dated 20 October 2026, reported 13,940 candidates and a 47% pass rate. Each percentage belongs to its own level’s sitting population.
| Level I | 28,145 | 43% | 24 September 2026 |
| Level II | 10,004 | 45% | 30 September 2026 |
| Level III | 13,940 | 47% | 20 October 2026 |
Do not add these percentages or treat the three groups as a single candidate cohort progressing through the programme. The people who sat Level III in August were not the same group who sat Level I in August. The denominator also matters: the release counts candidates who sat the exam at that level, not everyone who registered, studied or considered registration. The published percentage is rounded, so multiplying it by the candidate count can at most produce an approximate number of passes, not an exact audited headcount. The clean description is “CFA Institute reported that 47% of candidates who sat August 2026 Level III passed,” not “47% of CFA candidates passed.”
What a pass rate can and cannot say
A pass rate is a group outcome for a dated administration. It can help a reader understand the scale of that administration and compare published windows cautiously. It cannot establish that a particular question was harder, that one preparation method caused a change, or that the minimum passing score was equal to the pass rate. The rate is a share of candidates; the minimum passing score is a scoring threshold on an exam scale. They are different quantities with different denominators and purposes.
CFA Institute’s exam results and pass rates page explains that the minimum passing score may be set after an administration when there is a significant change in curriculum or candidate population, or maintained through equating when there is no significant change. Equating aims to account for differences in test difficulty across administrations. That official explanation is why a lower published pass rate should not be translated directly into a claim that the exam’s passing standard moved. The article’s table therefore reports outcomes only; it makes no estimate of a hidden score threshold.
The same page says a candidate receives an official pass or fail result with information about overall scaled performance and topic-level performance, subject to an important exception: a candidate who passes Level III does not receive detailed results. The public pass-rate release is not the individual result report. If you are interpreting your own performance, use the report that CFA Institute makes available to you, and apply its own explanation of the scale and minimum passing score. A public percentage cannot diagnose which topic cost you marks.
Why first-time and deferred groups need care
CFA Institute also reported different pass rates for candidates who were testing for the first time and those with at least one deferral. For August 2026 Level I, those reported group rates were 49% and 29%; for Level II, 52% and 30%; for Level III, 56% and 31%. Those are descriptions of groups in the same window, as published in the level-specific releases. They do not prove that deferring itself caused a lower result. The groups may differ in prior circumstances, preparation time, reasons for postponement and other unreported ways.
A candidate deciding whether to defer should therefore avoid treating a reported group gap as a personal forecast. The releases do not give enough information to isolate the effect of the scheduling decision. A more useful question is what would change in the candidate’s actual preparation before the new date: coverage of weak areas, quality of timed practice, available study hours and whether the revised plan is feasible. None of that can be inferred from the population percentages alone. The group result is a prompt to examine a plan, not a calculation of personal odds.
A worked interpretation without invented performance claims
Imagine that a colleague sees the August Level III rate of 47% and says, “I need to score at least 47% to pass.” The first correction is the unit: 47% is the fraction of candidates who passed, not the fraction of questions a candidate had to answer correctly. The second is the source of information: CFA Institute’s results page discusses a minimum passing score and equating, but the press release does not publish a personal cut score. The third is timing: the 47% figure refers to August 2026 Level III, not the colleague’s future window. The sound conclusion is only that CFA Institute reported 47% passing among that window’s Level III sitters.
Now suppose another colleague treats the three rates as proof that Level III is easier than Level I because 47 exceeds 43. That comparison ignores who was eligible and who sat each level. The candidate pools have different histories, and Level III candidates have already cleared earlier stages. The exam formats and content also differ. The published numbers permit a descriptive statement about each sitting, but not a causal ranking of inherent level difficulty. Before comparing two percentages, ask whether they have the same population, definition and exam window. Here they do not share the same population.
Use the result for the decision in front of you
A candidate who passed should follow the level-specific next step from CFA Institute rather than infer it from the public percentage. The August 2026 Level I and II announcements say successful candidates may progress to the next level. The Level III announcement says successful candidates can become charterholders only after work-experience requirements for charterholder membership are verified; charterholder members must also attest annually to the Code of Ethics and Standards of Professional Conduct. Passing the final exam is therefore a major exam milestone, but the release does not say every successful sitter automatically becomes a charterholder. The CFA programme page is a route into the wider study context.
A candidate who did not pass has a different immediate question: which parts of the personal result report indicate relative weakness, and which study decisions can be changed? CFA Institute says detailed scores and responses to exam questions are not released to candidates. That limits how precisely any candidate can reconstruct lost marks. It still leaves room to examine the available topic-level information, identify recurring errors in independent practice and plan a more targeted revision cycle. That public statistic offers no such diagnosis. The Omni method page describes an approach to structured practice; it is not a claim that a method guarantees a result.
Results are also time-sensitive data. The August 2026 Level III announcement was the latest of the three August releases as of this review on 25 October 2026, while other exam windows will eventually supersede it. CFA Institute’s results page is the continuing source for official releases and historical rates. Any later reader should check its current window before using these figures for a new comparison. That date check is especially important when an article is found through search long after its preparation.
A disciplined reading of the headline
Keep four labels attached to every published exam statistic: issuing body, level, exam window and population measured. For this article, that means CFA Institute, the named level, August 2026 and the candidates who sat. Then separate the group rate from the private result and the score threshold. These simple distinctions answer the useful question in the headline while avoiding a prediction that the data cannot support. The releases document what happened in a completed sitting; preparation choices still require an assessment of the candidate’s own circumstances.
Sources
- CFA Institute reports results for CFA Program Level I testing in August 2026 · cfainstitute.org · Retrieved
- CFA Institute reports results for CFA Program Level II testing in August 2026 · cfainstitute.org · Retrieved
- CFA Institute Reports Results for CFA Program Level III Testing in August 2026 · cfainstitute.org · Retrieved
- CFA exam results and pass rates | CFA Institute · cfainstitute.org · Retrieved



