The Accelerants

The LSAT is timed on purpose. Speed and accuracy together are what it measures. Law school exams need both, so do bar questions, and so does a junior associate at 11 p.m. with a client call at 7 a.m. A lawyer who reasons well but slowly is worse than one who reasons well and fast, so the clock is part of the test.

Give someone 50% more time and you’ve built a different test. Questions a standard-time taker has to skip and guess on can now be worked all the way through. The time pressure that forces fast, accurate judgment is gone, which changes what the score measures.

This matters because 2 things over the past decade have pushed a lot of scores higher, and neither has anything to do with reasoning ability. The first is a big jump in extended-time accommodations (extra time given to test-takers with documented disabilities). The second is organized commercial cheating. Both raise the number without raising the skill, and the score a student submits doesn’t tell schools which kind of number it is.

Before I say anything else about accommodations, I want to be clear about what this chapter is arguing and what it isn’t.

Students with real disabilities deserve equal access to this exam, full stop. The accommodation system exists for a good reason, and people who use it honestly are not the problem. The target here is the system itself. The no-flag policy. The near-total approval rate. The setup that’s made a neuropsychological evaluation (a professional assessment to document a disability) a smart buy for well-off families. Those are the design choices worth looking at, not the students.

Granted extended-time requests rose from roughly 6,000 in 2018-19 to roughly 15,000 in 2022-23, a 2.5-times increase, according to LSAC’s own research report TR-24-01.¹ The earlier climb, in the broader category of all accommodation types, was just as steep: approved requests of all kinds ran about 729 in 2012-13 and roughly 3,000 by 2016-17, already a fourfold jump before the extended-time numbers above. The approval rate in 2022-23 was about 98%. For 2024-25 it was about 99%.² ADHD was the single largest category over 5 years, at about 37% of all requests.

The most common accommodation is 50% extra time. Standard LSAT sections run about 35 minutes. Extended-time takers get about 52.

Extended-time scorers average about 4 to 5 points higher than standard-time takers with otherwise similar scores.³ That’s roughly the gap between a 164 and a 169, which is about 2 schools and tens of thousands of dollars in scholarship money. Often it’s the difference between a school you’d say yes to and one you wouldn’t.

LSAC does not flag accommodated score reports. A 170 earned with extra time looks the same to admissions offices as a 170 earned in standard time, and schools have no way to know.

For a well-off family, the incentive is straightforward. A private neuropsychological evaluation, running roughly $2,000 to $5,000, produces documentation LSAC approves about 98% of the time. What follows is 4 to 5 extra points on test day and a scholarship that tracks the higher score, all of it legal, traceable, and likely worth tens of thousands of dollars, and none of it about ability.

I’m laying this out because the incentive exists and the system built it. The fault is the design, not the people who respond to it.

Now the data, and the caveat right behind it, because LSAC has done this research honestly.

LSAC’s own data finds that extended-time LSAT scores overpredict (predict better grades than students actually earn) first-year law school performance. The most-cited study compared 121,378 standard-time takers with 880 extended-time takers who had near-identical LSAT scores. The extended-time group’s first-year GPA ran about 0.43 standard deviations lower. The prediction index overpredicted with residuals of 3 to 4 points on a scale whose standard deviation is 10. An earlier study found roughly 20% of extended-time takers landed in the bottom decile (lowest 10%) of first-year GPA, compared to about 10% in the top fifth.⁴ LSAC has found the same pattern across multiple groups.

Now the caveat, stated plainly. These studies mix disability status with extra time. They don’t track whether the same students also got accommodations on their first-year law school exams. A student with ADHD who also got accommodation in law school might do differently than the studies suggest. The data shows overprediction as a statistical pattern, without telling us why the pattern exists for any one person. It doesn’t say most accommodated test-takers are gaming the system. It says the score, unflagged and as it stands now, is a less reliable predictor for this group than for standard-time takers. That’s a problem for schools trying to build a class.

The no-flag policy is where this turns. If LSAC flagged accommodated scores on the reports it sends schools, schools could factor that in. They could weight the LSAT differently, ask for more information, or read the application another way. Right now they can’t. LSAC’s position, as of mid-2026, is that flagging would stop eligible students from asking for accommodation.⁵ That’s a real concern. It’s also a policy choice with costs the current design doesn’t admit.

The second accelerant is simpler to describe and less ethically complicated.

When the LSAT went digital and test-takers could sit at home, organized commercial cheating took off. On August 18, 2025, LSAC announced it would suspend online LSAT testing in mainland China, citing organized misconduct. Services were charging up to $8,000 per sitting. About 500 candidates in mainland China tested in 2024-25.⁶

The mechanics aren’t complicated. A skilled proxy tester sits in front of the camera while the real candidate gets questions through another channel, or the proxy sits in the candidate’s place entirely. Sharing test questions in real time lets organized operations build answer keys mid-test and hand them to coordinated groups.

On February 11, 2026, LSAC went further. From August 2026 the LSAT will no longer be offered online at all. After the June 2026 sitting, every test is in person.⁷ A test that went digital partly for access and volume is going back to proctored testing centers because remote testing couldn’t be kept secure.

For the score pool, the effect is statistical. Detection systems built to catch huge, unbelievable score jumps miss moderate, believable inflation. A candidate coached to hit a 167 on a 163 baseline doesn’t look suspicious. Neither does a coordinated group all landing in the 166-169 range with no individual red flags. But a small share of fake scores stacked at the top of the distribution still moves the medians, and the medians decide scholarships. The cheaters raise the curve for everyone.

Put both accelerants together. A real accommodation system that’s grown 2.5 times in 5 years, carries a near-100% approval rate, produces a 4-to-5-point score lift, and leaves no trace on the reports schools receive. And organized cheating that got serious enough to end remote testing worldwide after one country was suspended.

Neither one accounts for the majority of scores. Honestly, most test-takers sit under standard conditions, prepare hard, and submit real numbers. But a meaningful share of scores are inflated, schools can’t tell which ones, and the pressure to chase the median means they pay for all of them the same.

Watch: The LSAT Extra Time Advantage No One Wants to Talk About

The LSAT Extra Time Advantage No One Wants to Talk About

This chapter is from Anyone Can Beat the Law School Collapse by Steve Schwartz, LSAT Coach and Founder of LSAT Unplugged. I've been coaching the LSAT since 2005.

Published July 28, 2026.