Predict your SAT score from any practice test
The best way to predict a Digital SAT score is a full-length official practice test in Bluebook, scored the way the real test is scored. This predictor lets you do the same thing with whatever you have: the number of questions you got right in each module of a practice test, a paper test you marked by hand, or a single section from a prep book. Enter the four module counts (or the two section totals) and you get a projected total, both section scores, a likely range and your percentile among test-takers.
Reading the prediction honestly
Treat the number as the centre of a range, not a promise. Two things move a real score away from a prediction. The first is the curve: College Board equates every test form, so the same raw score can be worth 10–20 points more or less depending on the date. The second is the adaptive route: your Module 1 performance decides whether you see the harder or easier Module 2, and only the harder module reaches the top of the scale. The predictor applies the same routing rule the official practice tests use, and it shows you which module it assumed next to each section score.
Turning a prediction into a plan
The useful part of a prediction is the gap between it and your target. Because the curve is steep near the top and gentler in the middle, the same number of extra correct answers is worth more at a 550 than at a 750. Use the sliders to test scenarios: how many more Math questions turn a 640 into a 700, or how much an extra Reading and Writing question is worth once you are above 700. Then read how to improve your SAT score for what actually closes the gap, and check what your target means with what is a good SAT score.
Which practice tests to use
- Bluebook full-length tests (free, official): the most accurate, because they use the real adaptive engine and retired questions.
- Official paper practice tests: real questions, but linear. Score them here rather than with the paper worksheet if you want a single number instead of a range.
- Third-party tests: fine for practice, weaker for prediction, since their raw-to-scaled curves are estimates of an estimate. Weight them less.
Last updated September 2026. Predictions are estimates modelled on official practice-test data; official scores come only from College Board.