Predicting your CASPA GPA before you submit

By the time CASPA tells you what your verified GPA is, your school list is already submitted, the fees are already paid, and the verification window has already started ticking. The applicants who don't get caught off guard are the ones who run the math themselves before they submit. This is what that math actually looks like and how to do it.

The CASPA GPA is the one that counts

Every PA program receives one set of GPAs from CASPA: cumulative, science, non-science, Last 60, post-bacc if applicable, and graduate if applicable. They do not receive your transcript-reported GPA. They receive what CASPA calculated. So when you're building a school list, screening for fit, or deciding whether to apply at all, the GPA you should be benchmarking against is the CASPA-verified one - not the one on your transcript.

The two numbers usually differ. Sometimes by 0.02. Sometimes by 0.15. The reasons they differ are the things covered in the science GPA piece and the quarter-conversion piece: BCPM reclassification, retake averaging, withdrawal accounting, quarter-credit conversion. Each of these is a small adjustment. Stacked together they move your GPA materially.

Why predicting beats waiting

The CASPA cycle costs about $189 for the first program and $59 for each additional. Most pre-PA applicants apply to 8-15 programs. The total is somewhere between $600 and $1,000 in CASPA fees alone, before supplemental fees, transcripts, and the rest. A school list built around the wrong GPA - one that includes too many programs above your verified GPA, or one that misses programs that were a fit at the actual GPA - costs you both money and a cycle.

The verification window is also a factor. CASPA can take 3-6 weeks to verify your application after submission. If your verified GPA comes back materially different than what you expected, you can't easily add or remove programs at that point. The school list locks the moment fees are paid.

How to do the math yourself

You can do this by hand. It takes about three hours per transcript if you're careful. The steps:

1. Pull every course off your transcript. Title, credits, grade, term.
2. Reclassify each course into a CASPA Subject Category. The categories are listed in CASPA's coursework section help docs. Use the actual title and the department, not just the title - a "Statistics for Behavioral Science" course is Math by title and Behavioral Science by department, and CASPA tends to side with the department.
3. Convert quarter credits to semester credits if applicable. Multiply by 0.667.
4. Apply retake averaging. Both attempts of any retaken course count.
5. Handle withdrawals. W's are 0/0. WF's count as F.
6. Drop AP, IB, and CLEP credits from GPA calculations (they have no grade).
7. Compute weighted averages across all qualifying credits for each GPA bucket.

That's the manual approach. It works but it's slow, the reclassification step is error-prone (about 5-10% of courses get reclassified differently than the applicant expected), and the math compounds - one row misclassified can move the science GPA 0.05.

How to do the math in five minutes

The Transcript Mapper does all seven steps automatically. You upload your transcripts as PDFs (up to four, for combined-school applicants), the mapper extracts every row with two-pass verification, classifies each course into its CASPA Subject Category, applies the conversion and averaging math, and emails you a 7-page branded PDF report in about five minutes. The report includes:

• Predicted CASPA GPAs across all six categories: Cumulative, Science, Non-Science, Last 60, Post-Bacc, Graduate.
• A row-by-row classification table showing every course, its CASPA Subject Category, the credit conversion if applicable, and any retake-averaging math.
• Reclassification-risk flags for courses where the classification is ambiguous (Statistics-as-Math vs. Statistics-as-Behavioral, Nutrition-as-Other-Science vs. Nutrition-as-Non-Science, etc.).
• Withdrawal-vs-WF flags for any W rows that look like they might get reclassified.
• An estimated cumulative confidence interval based on how many ambiguous classifications were flagged.

Beta release - what that means

The mapper now lives inside the Application Dashboard. The engine works - the first ground-truth case (a two-school applicant on combined semester + quarter credits, with three withdrawals, one retake, and post-bacc coursework) came back with a predicted Combined Science GPA of 3.564 against an actual CASPA-verified Science GPA of 3.56, a difference of 0.004. That's well inside the band of useful for school-list strategy. Accuracy is still being refined as more transcripts come through, which is what the beta phase is for. Your submission becomes part of the calibration set, and the model gets better with every transcript that runs through it.

Dashboard access starts at $48 for three months. Up to four transcripts per submission. Sample report on the mapper page.

What to do with the prediction

Once you have the predicted CASPA GPAs, the school-list strategy follows naturally. Cross-reference the predicted GPAs against the published matriculant medians for each program - those are in the Program Spreadsheet and the Chances Calculator. Programs where your predicted CASPA GPA is at or above the matriculant median are competitive. Programs where it's 0.1-0.3 below the median are stretches that can still work with strong PCE, GRE, or essays. Programs where it's 0.3+ below the median are unlikely to interview-invite at scale, and the school list should weight toward fit rather than reach there.

The point of doing this in advance is that the school list shifts. An applicant whose transcript GPA suggests 12 programs but whose CASPA-predicted GPA suggests 8 different programs is going to apply differently. The fees are the same. The expected interview-invite count goes up. The cycle reads as more competitive on the program side. That's the leverage.

What I didn't realize when I was applying

The first cycle I applied - the one where I picked a single program and didn't get in - I didn't know my CASPA-verified GPA was different from my transcript GPA until the cycle was already underway. By the time I had the verified number, the fee was paid and the application was already at the program. The second cycle I applied to three. The third cycle I applied to twenty-two, with a school list built around the verified number this time. That cycle worked. The applicants who skip the prediction step are the ones who learn this the same way I did - one cycle late.

Frequently asked

How accurate is the mapper right now?

The first ground-truth case calibrated to within 0.004 on the combined Science GPA. The pre-launch test suite of 14 cases passed. The Dashboard rollout is what builds the larger calibration set. The calibration target is agreement within 0.05 for typical cases and 0.10 for edge cases - the Dashboard rollout is what builds the evidence across transcript types (international transcripts, heavy AP/CLEP credit, complex post-bacc + graduate combined math). If your situation is unusual, the mapper flags it.

What's in the report?

A 7-page branded PDF. Cover summary with all six predicted GPAs. Row-by-row table per transcript with classifications, conversions, and retake math. Flag list with explanations. Suggested actions if the predicted GPA suggests a school-list adjustment. Sample on the mapper page.

Can I run the mapper twice on the same transcripts?

Each submission runs through the engine separately. If you change anything between runs (a new term posted, a corrected grade, an additional transcript), the result will reflect that. If the inputs are identical the outputs will be too - the math is deterministic.

What if my prediction comes back lower than I hoped?

Then the school list adjusts and the cycle is built around the right number. The alternative - submitting against the wrong number - is what costs people the cycle. A lower-than-expected prediction is information, not a sentence. The reapplicant guide covers what to do if the gap is large enough that you need to think about a different cycle plan.