Proof that the AP Physics C Cutoffs increased
Starting with the 2025 exam, the format of the two AP Physics C exams changed. This mainly included a lot of additional time (35 MCQ in 45 minutes to 40 MCQ in 80 minutes, and 3 FRQs in 45 minutes to 4 FRQs in 100 minutes). There were some additional changes, such as the removal of the 5th choice from multiple choice, and changes to the questions themselves.
Previously, the Physics C courses were known to only require a score in the mid to high 50s or low 60s to get a 5, and a score of around 36% usually yielded a 3. When the test format changed, some believed that the cutoffs would increase to require a score of around 75% for a 5, similar to other AP courses, such as AP Physics 1 and 2. Others believed that the cutoffs would remain as they always were. However, it is clear that the cutoffs did increase, due to the following reasons.
- Evidence from AP graders.
This post contains the cutoffs obtained from AP graders. This information was shared by Foster Wisusik, who has been an AP Physics teacher and grader for a number of years. Essentially, it proves that for AP Physics C Mechanics and E&M, the cutoffs for a 5 were 76% and 81% respectively. He commented the following on the post:
“This is literally a real document the AP Teachers made using real data from last year (I am the person the poster is referring to and I was arguing with people in my livestream about the cut scores since they are led astray by Albert Io which was based on the old exam format). I am just surprised this is not widely known public information.
Here is how the teachers found this out and reverse-engineered the curve:
AP gives each teacher an average by Form, Grade, and AP Score. For example, I could see that my class averaged around a 4 on E&M last year.
If you are clever with the filters (sorting by score or grade level), you can end up finding the average of one student, and therefore isolate their score. For example, let's say you have one junior that got a 5 on Form 2. By applying filters (Junior, Took Form 2, Scored a 5) It told me the average of that group (one student) was 65/80. Thus, 65/80 is a 5.
For example, if you are the only 10th grader in your class taking your exam, your teacher can determine your exact percentage using these filters.
We sent this out to all teachers around the country, and they also did this process. Eventually, we got enough singletons to isolate the boundaries, and thus the cut scores were determined for all the courses. This project has been going on for many years. In 2024, the cut was much lower”.
There has already been a lot of discussion regarding this, so I will leave it at that, and spend the rest of the time providing additional evidence and discussing counterarguments.
2. “Even though the exam’s format changed, there is no way an exam with Maxwell’s equations in integral form would require an 80% for a 5”.
(First of all, E&M barely even covers Maxwell’s equations, but I won’t focus on that).
This is a claim that some use to dispute the fact that the cutoffs changed. Based on this claim, we can say that the cutoffs used on the AP exam are based on the difficulty of the content, and are independent of the exam’s format. However, this can be disproven by looking at other courses.
For example, in this released AP Bio exam from 1999, only a 57.3% was required for a 5 (page 100). However, on this practice exam from 2013, a 77.5% was required for a 5 (page 93). Is this because Biology magically got harder in 13 years? No. Instead, the main reason for the change in the cutoffs was due to differences in the exam’s format; the content stayed relatively similar, but the exam’s format greatly changed, causing different cutoffs.
For another example, consider the algebra-based physics courses. On this released exam from 1998, a 60% was required for a 5 on AP Physics B (page 224). Since then, AP Physics B split into AP Physics 1 and 2, and the format of the course and exam shifted (though the physics stayed mostly the same). However, on the 2018 AP Physics 1 practice exam, a 76% was needed for a 5, and on the 2019 AP Physics 1 practice exam, a 71% was needed (these are private documents, so I don’t want to link them, but you should be able to find them online). Again, this proves that changing the format of an exam can change the cutoffs.
3. If the cutoffs did not change when the exam’s format changed, that means that the average score should have greatly increased, based on publicly released FRQ data, but that did not occur.
(This is the link that contains the data I’m using).
On the 2025 exam, the average score on the Mechanics FRQ was 55%. On E&M, it was 60%.
From 2022-2024, the average score on the Mechanics FRQ was 45%. On E&M, it was 41%. (Note that each year had 2 publicly released sets).
Based on this information, if the cutoffs did not change, the average AP scores for these classes should have gone up. However, for both classes, it went down. For Mechanics, it was 3.3 in 2025 vs. 3.43 average in 2022-2024, and in E&M, it was 3.38 in 2025 vs. 3.49 average in 2022-2024. Because the average stayed mostly the same while the FRQ averages greatly increased, that would indicate that the cutoffs were probably adjusted in order to keep similar AP scores on an easier exam.
4. “AP readers are just random people, and CollegeBoard is the only one with the actual data”.
As Foster Wisusik explained, the process that the AP teachers used is very methodical based on the data they were given as is as close to the real thing as possible, and they noticed a huge increase in the cutoffs in 2025 due to the new exam format. So while CollegeBoard does try to be secretive about the cutoffs, they can easily be reverse-engineered.
Additionally, if this was true, then how would albert.io be accurate? Albert.io would be just as accurate as the useless AP readers, so if the AP readers are inaccurate, then albert.io wouldn’t be any better.
5. “There is no way AP Physics 1 and 2 have more lenient curves than the AP Physics C courses”.
Go back to the 1998 released Physics exams, linked here. As mentioned earlier, page 224 indicates that a 60% was needed for a 5. However, page 225 indicates that a 61% was needed for a 5 on AP Physics C: Mechanics, and page 226 indicates that a 55% was needed for a 5 on E&M. The exam format of the 3 exams was essentially the same, and the fact that the cutoffs were pretty similar indicates that adding calculus will not automatically decrease the cutoffs. With that logic, it isn’t unreasonable to assume that the curves for AP Physics 1 and 2 could be similar to the curves for AP Physics C: Mechanics and E&M.
6. “Why would the CollegeBoard increase the cutoffs for AP Physics C but keep classes such as AP Calc BC the same?”
As a reminder, just because the cutoffs for AP Physics C increased does not mean that the exam got harder, as the cutoffs only increased because the exam’s new format makes it easier. Because of this, this argument has logical fallacies.
7. “AP graders are exaggerating to overprepare you.”
I don’t have good evidence to dispute this, but when it’s the day before the exam, there’s not much extra preparation you could do. So when AP readers share this information at that point, it wouldn’t change how you prepare, as there’s not much you could do at that point, but instead would make you be mentally ready for what’s to come. Again, I don't have definitive evidence against this claim, but it is highly unlikely that the hundreds of AP teachers who participated in this progress in reverse-engineering the cutoffs collectively agreed to exaggerate them.
Final notes
- Remember that the cutoffs always slightly change year to year due to slight variations in exam difficulty, so just because the cutoffs for Mechanics and E&M were 76% and 81% respectively doesn’t mean that they can’t change this year. For Mechanics specifically (I didn’t take E&M), the 2026 FRQs were definitely harder than the 2025 ones, but the 2026 mcq was rather easy, but because the multiple choice questions aren’t publicly released, it is impossible to determine if the cutoffs will change.
- Because someone will probably mention it, remember that AP exams are not “curved”, but rather “equated” by comparing how high schoolers do to how college students perform. So the cutoffs are more based on the objective difficulty of the test rather than how well AP students perform.
- Here is a comment from Foster Wisusik. The context was someone mentioned that someone else said that the methodology was inaccurate due to there being multiple forms. "The scores are filtered by form. Each one has a separate required score to account for differences in difficulty. Overall, the test got way easier with the 2025 redesign due to 2x time and removing an answer choice. The methodology is absolutely correct and the listed cut scores are valid for Form J of the 2025 exams last year. I’m sure it will change this year, but we will easily deconstruct the cuts again. Usually only the US form J gets figured out completely. There are a bunch of ignorant people online that refuse to believe the cut scores changed because of an Albert IO calculator from the old format. I’ve given up trying to convince them. Let them be wrong”.
And finally, if you have any other points you would like to bring up, comment on them, and I’ll respond to them.
Edit: as some people (including myself in this post) have correctly pointed out in the comments, score cutoffs change from year to year due to slight changes in exam difficulty. However, it is hard to imagine that that slight variation is the reason why the cutoffs were so high in 2025 compared to beforehand, and it is likely that the cutoffs will permanently be around 75% as opposed to the old 60%.
For evidence of this, the highest cutoff I’ve seen on a test from the old format is a 70% for a 5, which is from the 2013 AP Physics C: Mechanics Practice Test (it’s not supposed to be publicly accessible, so I won’t link it, but you can easily find it online). The reason for the high cutoff was because this test was unusually easy (I took it myself and got a much higher raw score than on any other old-format practice test). If the high cutoffs in 2025 were due to normal variation in difficulty, then they should not have risen much higher than 70%. Instead, it is likely that the cutoff was caused by the new format, and that future cutoffs probably won’t vary too much from a score in the 70s needed for a 5.