Stop Asking ChatGPT What Will Be on the Bar Exam

Right now, days before the July 28-29 bar exam, thousands of people are typing some version of "what will be on the bar exam" into ChatGPT. It's a completely understandable thing to do when you're anxious and the clock is running. It's also a waste of the little time you have left.

Here's the truth nobody selling you a prediction wants to say plainly: nobody has a crystal ball. Not ChatGPT, not Claude, not the prep companies, and not this blog. Bar examiners do not publish their topics in advance. So every confident forecast you read is a guess wearing a suit.

That's not a reason to panic. It's a reason to point your last six days at things that actually move your score. Let's do that.

Can ChatGPT predict what's on the bar exam?

When you ask a chatbot what will be on the bar exam, it doesn't know anything you don't. It pattern-matches against old exams it has seen and produces a tidy, confident-sounding list. Confidence is not accuracy. The model has no inside information, because none exists to have. That's not a knock on the tools. They're genuinely useful for explaining a rule or talking through a fact pattern. Predicting a secret topic list is simply not a thing they can do, because the data doesn't exist.

Try a quick experiment. Ask the same question twice, in two fresh chats. You'll often get two different lists. That alone should tell you what's going on. A tool that changes its "prediction" based on how you phrased the prompt isn't predicting. It's generating plausible text.

The real danger isn't that the guess is wrong. It's what a wrong guess does to your week. If a chatbot convinces you that, say, Evidence is "unlikely" this time, you might quietly stop reviewing it. Then Evidence shows up, and you've handed away points you could have earned. Chasing predictions starves the subjects that actually appear. With six days left, you can't afford a single hour spent studying a forecast instead of the exam.

What was actually tested on the February 2026 bar exam?

The most recent administration was February 2026. Here's what got tested on the MEE, as reported by the prep provider JD Advising in their post-exam recap. The NCBE itself does not pre-announce or, at the time of writing, freely publish the topic list, so read this as a provider's recap rather than an official source:

• Secured Transactions (attachment, perfection, sale to a buyer not in the ordinary course)
• Real Property (duty to mitigate, co-tenant profit sharing, ouster, severance of a joint tenancy)
• Corporations (which crossed over into Agency and Partnership: incorporation defects, partner liability, shareholder liability, preincorporation contracts)
• Evidence (relevancy, character evidence, MIMIC)
• Civil Procedure (service of process, notice, personal jurisdiction)
• Decedents' Estates (antilapse, ademption, residuary estates, pretermitted children)

Here's what that list does and does not tell you. It does tell you that crossovers are real: the Corporations essay blended in Partnership issues. It does not tell you what comes next week. In fact, two of those subjects, Secured Transactions and Trusts and Estates, leave the MEE starting with the July 2026 exam, so February 2026 was their last standalone appearance. Looking at that list and reverse-engineering July's exam is exactly the trap. Use it to understand the shape of the test, not to place bets.

And July 2025 before that

Go back one more administration and the point gets clearer. The July 2025 MEE subjects, confirmed from the official NCBE question set (each question prints its subject on the cover page), were:

• Corporations and LLCs
• Contracts
• Trusts and Future Interests
• Constitutional Law
• Criminal Law and Procedure (the tested content was Criminal Procedure: Fourth Amendment searches, including school searches)
• Torts (negligence, vicarious liability, NIED)

Line up July 2025 against February 2026 and notice how little overlap there is in some spots and how much repeats in others. There's no clean rotation you can decode. If there were a pattern reliable enough to predict, the prep companies would sell it as a guarantee instead of a heavily hedged "educated guess."

What is bar exam history actually good for?

History isn't useless. It's just useless for prediction. What it's genuinely good for is prioritization: deciding where to spend your limited hours when you can't give every subject equal attention.

Frequency charts from prep providers (and the analyses that agree with them) point in a consistent direction. Civil Procedure is the single most frequently tested MEE subject and shows up on nearly every administration. Contracts, Real Property, Business Associations, and Evidence all rank high historically. Those are the subjects that reward review time most, because they appear again and again.

Now, the myth that trips people up. You'll hear that a subject "won't repeat because it was just tested." That's mostly false, and the recent record proves it. Per JD Advising's own materials:

• Constitutional Law was tested three administrations in a row: July 2024, February 2025, and July 2025.
• Torts was tested back to back in February 2025 and July 2025.
• Corporations and LLCs appeared in July 2024 and again July 2025, and earlier ran three straight administrations.
• Civil Procedure shows up so often it effectively repeats every time.

So do not cross a subject off your list just because it appeared last time. High-frequency subjects repeat constantly. The prep companies themselves say it best: frequency charts are prioritization tools, not prediction tools. Use them to rank your review, not to gamble on what's "due."

The only plan that works with 6 days left

You can't learn everything between now and Tuesday. You can walk in sharp, calm, and rehearsed. Here's a day-by-day plan for July 22 through 27 built for exactly that.

Tuesday, July 22 (today). List the high-frequency subjects, then mark the ones you feel shakiest on. That intersection, the subjects that are both commonly tested and weak for you, is your target for the week. Do a focused review pass on the first one. Then write one timed essay in that subject and get real feedback on it.

This is where practicing the way you'll be graded matters more than another lecture. It's not enough to know a rule. On test day you have to write it into a point-earning answer under pressure. This is exactly where BarScore fits: paste your essay and prompt, and in seconds you get a score plus specific feedback on the rules you missed, where your analysis went thin, and what would earn more points. That beats staring at a model answer and guessing why yours fell short. It works with any prep course, MEE or state essays, in any US jurisdiction, and there's a free trial on its subscriptions if you want to run your practice essays through it this week.

Wednesday, July 23. Second weak high-frequency subject. One timed essay, feedback, revise. Keep your MBE muscle warm with a short set of 25 to 35 questions and review every miss. Short and daily beats one exhausting marathon.

Thursday, July 24. Third target subject, same rhythm: review, one timed essay, feedback. Another small MBE set. Notice you're doing one graded essay a day, not ten. Quality of feedback beats raw volume this late.

Friday, July 25. Finish your logistics today so the weekend is clean. Confirm your test center and start time, print or load your admission ticket, lay out your ID and any approved supplies, plan your route and parking, and check what you're allowed to bring. Then one more essay and an MBE set. Getting logistics done now removes a whole category of test-week panic.

Saturday, July 26. Light, targeted review of your highest-yield subjects. One timed essay if it helps your confidence, but don't grind. Reread your own best answers from the week so the structure that scores well is fresh in your mind.

Sunday, July 27. Taper. Skim your outlines, review the feedback themes that kept coming up in your practice essays, and stop early. No new material. Your job tonight is to protect your sleep, not to add facts. Cramming Sunday night trades a rested brain for a few flashcards, and that's a bad trade.

Then Monday, rest and arrive ready. You've prepared for the exam itself, not for a prediction.

Skip the fortune-telling. Point these last days at the subjects history says matter most and at writing essays that actually earn points. If you want a fast read on where your practice essays stand this week, that's what BarScore's feedback is for.