What predicts success in law school? The standard answer is your LSAT and GPA. Each correlates with first-year grades, and together they predict better than either one alone.
The numbers themselves are not the cause, though. Earning a high LSAT score takes one skill, and earning a high GPA takes a different skill. Those two skills are what carry into law school and drive first-year grades.
There is also a third skill that matters as much as either, and no part of the application measures it.
What the Numbers Are Reading
Start with what each number is actually a measurement of because the two are not measuring the same skill.
What a GPA Is Reading
Four years of grades measure whether you can work. Show up, put in hours, study without supervision, sustain it across semesters when nothing forces you to.
They also measure something less obvious: whether you can figure out what a course rewards and deliver that. Every professor wants something slightly different, and strong students diagnose it early and adjust. That diagnostic habit is doing a lot of the work behind a high GPA.
What an LSAT Score Is Reading
The LSAT does two things at once: it tests whether you can train your brain to think a certain way, and the prep itself is that brain training. The studying builds the skill, and the score reports whether it got built. Most people do not start out thinking the way the test demands. Months of drilling build it, until the moves fire on their own.
That resolves two questions applicants ask about the test:
→ Your highest score is the real score. A 172 after multiple takes is a brain trained to reach 172. Averaging the attempts would measure where you started rather than what you built, which defeats the point of measuring.
→ Take count is noise. Test-day variance, one bad section, being ready in June instead of January: none of it changes what the final number certifies.
More on the retake math in Yes, You Should (Probably) Retake the LSAT.
The Two Numbers Are Not Equally Reliable
Both skills are real. The numbers reporting them went through very different processes:
| Reliability | Recency | |
|---|---|---|
| LSAT | Same test, same scale, everyone | Months before you enroll |
| GPA | Grade inflation, your school’s grading culture, your major’s difficulty | However long ago college was |
The reliability gap shows up in the data. LSAC’s 2019 correlation study put the LSAT’s correlation with first-year grades at about .40 against GPA’s .26, and about .60 against .42 when normalized. And LSAC’s 2021-2025 summary reports the gap widening: grade inflation has pushed so many applicants toward 4.0 that UGPA has a ceiling effect, and it has become a weaker predictor as a result.
Recency is the practical gap. Most applicants’ LSAT is recent. A GPA can be current for the senior applying straight through, or years old for the career changer. The older it is, the less it says, because the work engine can be rebuilt after the number locks.
A weak GPA from years ago is a weak predictor of law school grades.
Law School Asks for Both Skills Again
1L is a second round of both. The brain-training skill gets a new target, and the work engine gets a new game to figure out.
The LSAT pays you to collapse ambiguity fast: one credited answer, wrong choices built to be eliminated. A law school exam still wants a conclusion, and speed still matters. But most of the points sit earlier, in the elements the fact pattern made genuinely arguable.
The new target for your thinking runs toward ambiguity instead of away from it.
That is not bad news for high scorers. You already proved you can train your brain to a target. Law school just hands you a new target: this time, the training is learning to hold the question open before you close it.
An answer that states a conclusion and moves on has caught the bouncing ball on its first bounce, the scoring model at the center of my law school exam guide: the points live in the bouncing, not the landing. You still conclude. You conclude last.
The work engine’s new game is subtler. The diagnostic habit behind your GPA, figuring out what a course rewards, is still the right habit. But the thing to diagnose changed: what a law school exam rewards is not stated in the reading or on the syllabus. And the feedback that powered the diagnosing is mostly gone. In college, constant grading told you whether your approach was working. In 1L, the grade usually arrives once, at the end, after it counts.
The effort and the diagnostic habit transfer. The feedback that steered them disappears.
Which is why the students who keep their college trajectory manufacture the feedback themselves:
→ Start your professor’s old exams early in the semester. They show you what the exam actually demands while there is still time to adjust how you study.
→ Bring written answers to office hours. Your professor tells you where the analysis falls short.
More first-year habits in 10 Key Lessons to Succeed in Law School.
The Skill Your Application Never Measured
Everyone in your section cleared roughly the same numerical bar, so inside the room the numbers stop doing the separating. What separates people starts before any exam, in how they read.
In a study published in the Seattle University Law Review, Professor Leah Christensen had first-year law students read the same judicial opinion and compared how the top and bottom halves of the class went about it. The halves read in fundamentally different ways. Call the two modes absorbing and reconstructing:
The bottom half read to absorb: find the rule, mark the important-sounding language, move on.
The top half read to reconstruct the reasoning: why the case comes out this way, what is actually driving the outcome, how the court gets from these facts to this result.
Take International Shoe Co. v. Washington, the personal jurisdiction case every 1L reads and courts still apply eighty years later. A Delaware company with no offices in Washington has commissioned salesmen working the state, and Washington wants to haul it into its courts.
- The absorb reader leaves with the rule. A state can reach a defendant with minimum contacts, so long as the suit does not offend fair play and substantial justice.
- The reconstruct reader leaves with why. The old rule tied jurisdiction to physical presence, and a corporation selling through traveling salesmen is everywhere and nowhere under it. So the Court swaps where the company is for what it does there: a company that works a state’s market and enjoys its protection can answer in that state’s courts for that business.
Rebuild that, and “minimum contacts” stops being a vague phrase. It is a reciprocity line. And the reader who rebuilt it can argue the next fact pattern, the website order, the single shipment, in both directions.
Reconstruction is not a case brief. A brief records what the opinion says, and you can fill one out on autopilot. Reconstruction means the reasoning is assembled in your head, so you could rebuild the decision yourself.
The test for real understanding: Can you explain the case to someone who has never read it? If not, you do not understand it yet.
This is why highlighting fails. It is a passive activity disguised as an active one: the marks make the pass feel like work, but nothing is being assembled.
Final Word
Effort without trained reasoning stalls in 1L. Trained reasoning without reconstruction produces confident, shallow exam answers. Reconstruction without the discipline to do it in five subjects is a skill that never gets exercised.
Your file reported two of the three as numbers. All three keep moving after the numbers are set, which is why a number is a reading and not a forecast.
Applying this cycle?
I work with applicants one-on-one on the whole application process, from strategy through essays and submission.
Explore my services | Request a consultation
Related Reading
→ How to Succeed on Law School Exams: A Full Guide
→ 10 Key Lessons to Succeed in Law School
→ My 1L Civil Procedure Exam
→ Yes, You Should (Probably) Retake the LSAT
→ Blog Directory