CFA Score Report Explained: What 1600, Your Score Box and "Above 70%" Actually Mean

The CFA score report is built to tell you a great deal about your performance while revealing almost nothing about the exam. That tension is why it looks cryptic on first reading. It is not, once you know what each element represents.

The report also changed recently, which is why a large share of the guides you will find online describe percentile lines that no longer exist. This covers the current format, element by element, and then the part most guides skip: what to actually do with it.

The headline change: scale scores

From the February 2025 administration, CFA Institute began reporting a scale score at all three levels. A scale score is your raw score — the number of questions answered correctly — mathematically transformed onto a fixed scale so results are comparable across different forms of the same exam.

• Level 1: 1000 to 1900, minimum passing score 1600
• Level 2: 2000 to 2900, minimum passing score 2600
• Level 3: 3000 to 3900, minimum passing score 3600

The leading digit only identifies the level. Strip it and every level is scored 0 to 900 with 600 as the bar.

Why a scale score is not a percentage

This is the single most important thing to understand about the format, and it is where most candidate confusion lives.

The MPS is a fixed point on the reported scale, but it is not a fixed number of correct answers. For each administration, a panel of charterholders sets the standard using a modified Angoff process, judging how a minimally competent candidate would perform on each question. That produces a raw threshold specific to that exam form. Equating then maps that raw threshold to 1600, so a candidate who sat a harder form is not penalised and one who sat an easier form gains no advantage.

The consequence: a scale score of 1650 does not mean 65%, or 82%, or any other percentage. There is no conversion table. Any site offering one is guessing, and the "you need 70% to pass" rule circulating in forums is folklore rather than policy — a useful study target, but not a statement about how grading works. See the MPS explainer for how the standard-setting works in detail.

What a given scale score tells you

What the number does give you, which the old report did not, is a clear sense of margin. At Level 1:

• Under about 1450: a coverage problem. Substantial parts of the curriculum were not learned to exam standard.
• Roughly 1500–1590: a marginal fail. The knowledge base is largely there; the gap is usually two or three weak topics plus exam technique.
• 1600–1650: a marginal pass. You passed on this form and might not have on another.
• Above about 1700: a clear pass with real margin.

These bands are interpretive rather than official, but the distinction between the first two matters enormously for a retake, because they call for completely different plans.

What was removed, and why

Older reports plotted a 10th percentile and a 90th percentile line so you could see where you sat relative to the global cohort. Those are gone. CFA Institute's stated reasoning is that scale scores add precision to results interpretation, whereas the percentiles only provided a comparison against other candidates.

That reasoning reflects something important about the exam: the CFA is not curved. You are measured against a standard, not against the person next to you. The number of candidates in your window has no bearing on how many can pass. Removing the percentiles made the report honest about that.

From the August 2025 administration, group average performance for each topic area was added in graphical form. So cohort comparison still exists — but at topic level, where it is diagnostically useful, rather than at total level, where it was mostly a source of anxiety.

The score line and the confidence interval box

Your result appears as a line inside a shaded box. The line is your score. The box is a confidence interval: an estimate of the range your score would plausibly have fallen into had you sat a different form of the same exam, with a different draw of questions and a different run of lucky and unlucky guesses.

Read the box, not just the line:

• Pass, box entirely above the MPS: a robust result. You would have passed most forms.
• Pass, box straddling the MPS: you passed on a form that suited you. If you are moving to the next level, do not read this as evidence that your study method is sufficient.
• Fail, box straddling the MPS: genuinely marginal. Modest targeted work should be enough.
• Fail, box entirely below: the gap is structural, not circumstantial. Plan a full retake cycle rather than a top-up.

The topic breakdown, and the question nobody answers

Below the total, performance is shown by topic against two reference lines at 50% and 70% of available points. CFA Institute treats 70% as the marker for sufficient command of a topic.

Which raises the question candidates ask constantly and almost no site addresses: how can most of my topics sit above 70% while my total still fell short?

There are three reasons, and they are worth understanding before you plan a retake.

The bands are wide. The chart shows a band, not a point. A topic sitting "above 70%" may be at 71% with a confidence interval running from 60% to 82%. Several topics landing just over the line can easily sum to a total under the standard.

Topics are not equally weighted. Ethics carries far more questions than Derivatives. Being above 70% on Derivatives and below on Ethics and Financial Statement Analysis is a much worse position than the visual suggests, because the chart gives every topic the same amount of vertical space regardless of how many questions sit behind it.

You cannot reconstruct the total from the parts. CFA Institute is explicit that your score is not and cannot be calculated from the topic information provided. Topics contain different numbers of questions, so averaging the bands produces a number with no meaning. If your mental model is "my topics averaged 70% so I should have passed", the model itself is wrong.

Note also that topic confidence intervals are wider than the total's, because each topic contains far fewer questions. A single topic band just under 70% is weak evidence on its own. The same topic under 70% across two mocks and the real exam is strong evidence.

What the score report will never tell you

• Your raw percentage, or the number of questions you answered correctly
• Which specific questions you missed
• The MPS in raw-score terms for your form
• Whether the ethics adjustment was applied to your script

Results are also final. Multiple-choice and item-set questions are machine-scored, Level 3 constructed responses are marked by qualified charterholders, and retabulations are not available on request. Candidates used to university appeals processes occasionally find this surprising.

Turning a fail into a retake plan

If you did not pass, the report contains four usable signals, in descending order of reliability.

1. Distance from the MPS. This determines the shape of the plan. A high-1500s Level 1 score is a tuning problem — three or four weeks of targeted work on identified gaps plus mock discipline. A low-1300s score is a coverage problem and needs most of a full study cycle. Confusing the two is the most common retake mistake: candidates with coverage problems do a light refresh and fail again.

2. Topics consistently under 50%. These are knowledge gaps rather than technique gaps. They need re-teaching from the curriculum, not more practice questions. Doing question banks on material you never learned properly produces frustration, not progress.

3. High-weight topics in the 50–70% band. This is where the fastest marginal gains sit. You already have partial command, and the topic carries enough questions for improvement to move the total. Ethics, Financial Statement Analysis and Equity at Level 1 are the ones worth checking first — the topic weights tell you how much each band is actually worth.

4. The group average comparison. Where you sit well below the cohort average on a topic, that is a signal the material is learnable and you did not learn it — as opposed to a topic where everyone scored poorly, which usually means a hard question set.

The signal the report cannot give you

Timing. If you ran out of time and guessed the last twenty questions, the topic bands will look like knowledge gaps when the real failure was pacing. Nothing in the report distinguishes the two.

Only your own recollection can, and memory of the sitting decays fast. Write down, within twenty-four hours of the exam: where you were at the halfway point, whether you rushed the final stretch, how many questions you flagged and never returned to, and which topics felt slow. Eight weeks later that note will be worth more than the score report itself.

If you passed

Read the report anyway. Topics under 70% at Level 1 do not disappear — Level 2 assumes fluency in exactly that material and does not re-teach it. A candidate who scraped through with weak Financial Statement Analysis is walking into a Level 2 exam that leans on it heavily. Two weeks of remedial work before starting Level 2 is far cheaper than discovering the gap in month three. The Level 2 transition guide covers what changes and what is assumed.