How Many Practice Exams Should You Take?

For a standard four-credit doctrinal course with a three- or four-hour final, the working target is three to five full-length, timed, closed-laptop-conditions practice exams, taken in the last three to four weeks. Below three, you have not stress-tested your outline under time pressure. Above five or six, you usually hit diminishing returns — the gains stop coming from more reps and start coming from more careful diagnosis of the reps you already ran.

Layer shorter work underneath that. Eight to fifteen single-issue hypotheticals, old exam sub-questions, or Examples & Explanations problems per course, spread across the semester, do a different job: they teach you the rule structure while you still have time to fix it. A full timed exam in week six is mostly a demoralizing waste, because you have not seen two-thirds of the material.

The number genuinely varies, and three things drive the variation. First, format: a course with 60 multiple-choice questions demands high volume (100–200 practice MCQs) rather than four essays. Second, supply: some professors post ten years of exams with model answers; some post nothing. Third, your baseline — if your first practice exam comes back clean and organized, do fewer and spend the time on rule memorization; if it is a shapeless mess, do more and do them earlier.

Rank your sources by proximity to the actual exam you are taking. Your professor's own past exams, with any released model or sample answers, are worth more than everything else combined, because they reveal question style, fact-pattern length, the ratio of doctrine to policy, and which corners of the course the professor actually cares about. Check the course page, the library's exam bank (most law libraries maintain one, often behind a student login), and simply ask — many professors will hand over an old exam if asked in office hours.

Next best are exams from other professors teaching the same course at your school, then commercial sources. Siegel's essay-and-multiple-choice series and the Glannon Guide series give you questions with explanations. Examples & Explanations volumes are problem sets, not exams, but they are excellent for the mid-semester layer. CALI lessons are free through your school and useful for rule mechanics. For Torts, Contracts, Property, Con Law, Crim, Civ Pro, and Evidence, released NCBE multiple-choice practice questions are well-drafted, though they test bar-level majority rules rather than your professor's idiosyncrasies.

Avoid burning your professor's most recent exam early. Save one, ideally the most recent, to take under full simulated conditions in the final week.

Weeks 1–6: no full exams. Work short hypotheticals from your casebook's notes and from an E&E as you finish each unit. The goal is to convert reading into a usable rule statement plus a sense of what facts make the rule contested.

Weeks 7–10: take one full past exam untimed, open-outline, as a diagnostic. You will be slow and you will miss issues on material not yet covered. That is the point — it shows you what an exam question in this course looks like while you can still restructure your outline around it.

Final three to four weeks: this is where the three to five timed exams live. Space them roughly every three or four days per course, and never take two full exams for the same course back to back. You need the interval to fix what the first one exposed.

If you are reading this two weeks out, compress rather than skip. Two timed exams per course with real diagnosis beats five rushed ones you never review.

Match the real conditions. Same time limit, same word or page cap, same open- or closed-book rule, same software (run your exam software's practice mode at least once so you are not learning it live). If the exam is closed-book, do not peek at your outline; the entire value is finding out what you cannot retrieve cold.

Budget time explicitly on the exam itself. A common allocation for an essay is about one-quarter to one-third of the allotted time spent reading the facts and building a short issue list, then the remainder writing. Write that plan on your scratch paper before you start.

Write full prose answers, not outlines, for at least three of the exams. Outlining answers is faster and tempting, but the skill you are actually being graded on — moving from rule to fact to counterargument in readable sentences under time pressure — only develops if you produce the sentences.

Build the key yourself before you look at anything else. Reread the fact pattern slowly and list every legally significant fact and the issue it triggers. Then compare that list to what you wrote. Score two separate things: issues spotted versus issues available, and depth of analysis on the issues you did spot. Most students who feel they 'ran out of time' actually spent too long on an easy issue and never reached three hard ones.

If a model answer exists, do not read it as a checklist of conclusions. Read it for structure — where the model pauses to argue both sides, where it disposes of an element in one line, where it flags an ambiguity in the facts rather than resolving it.

Use your study group as a grader swap. Three people writing the same exam and cross-reading produces a better issue list than any one of you generates alone, and reading someone else's weak analysis teaches you to see your own.

Then go to office hours with something small. Professors will rarely mark an eight-page answer, but most will read one page, or will react to your issue list and time allocation. Ask a specific question: 'I spent twelve minutes on the offer-and-acceptance analysis here — was that the right weight?'

Finally, keep a running error log per course. Missed issues, misstated elements, conclusions with no counterargument. Before the next practice exam, reread the log. That document, not the exam count, is what moves your grade.

For multiple-choice components, count questions, not exams. Aim for 100–200 practice questions per subject, done in blocks of 25–35 timed at roughly 1.8 minutes each, and review every question you got wrong and every one you guessed right. The review is where the learning is; the answering is just data collection.

For take-home or 24-to-48-hour exams, do fewer full simulations — one or two — but practice the specific failure mode, which is over-research and over-writing. Practice hitting a word cap. Practice cutting.

For heavily policy-weighted or theory courses, full timed reps matter less than pre-drafting. Build two or three flexible argument frameworks you can deploy against varied prompts, then test them against past prompts in twenty-minute outlining sprints rather than full write-outs.

For open-book exams, practice with the exact outline you will bring. If you cannot find a rule in your own document in under fifteen seconds, the document is too long, and only a timed practice run will reveal that.