Permit Test: How Many Questions Can You Miss?

The Number Is a Diagnostic, Not a Budget

Work out how many questions you can miss and you get a different number depending on where you live. Almost everybody then treats that number as an allowance to spend, as though the job is to stay under it.

It is more useful the other way round. Your miss allowance tells you how many topics you are permitted to be shaky on. Three spare answers means one weak area is the entire margin, and a single unlucky draw from that area ends the attempt. Ten spare answers leave more room for weak topics. Same effort, completely different study plan, and the only way to know which one you are in is to do the arithmetic before you start revising rather than after.

So this guide does the subtraction, then spends most of its length on the four situations where the subtraction quietly gives you the wrong number.

The Subtraction

Two figures decide it: how many questions are on the paper, and how many you must answer correctly. Subtract the second from the first. That is the whole calculation, and for most states it is exact.

Where a state publishes only a percentage, multiply and round up, because there is no partial correct answer and agencies do not award one. Where a state publishes a raw score directly, use the raw score and ignore the percentage entirely, including any percentage you worked out yourself. The raw number is the rule; the percentage is a description of it, and the two do not always land on the same integer.

One case worth naming: Maryland requires 22 correct out of 25, which is 88 percent, higher than any of the round figures people expect. Pennsylvania sets 15 of 18, roughly 83 percent, on the shortest paper in the data set. Neither of those is a number you would have guessed from a percentage.

• Find the number of questions on your state's paper.
• Find the number you must get right, as a raw score if the agency publishes one.
• If only a percentage is published, multiply it by the question count and round up.
• Subtract. What is left is the allowance.
• Then read the next four sections, because four things can make that number wrong.

SourcesMaryland Department of Transportation Motor Vehicle Administration, Knowledge Tests, 22 of 25 correct, about 88 percent (checked 2026-08-27); PennDOT, Pennsylvania Driver's Manual, Testing, 15 of 18 correct, about 83 percent (checked 2026-08-27)

Worked Example: Two Applicants, Both Told They Need 80 Percent

One is sitting the Kansas test, the other the Wisconsin test. Both have been told the pass mark is 80 percent, and both believe they are facing the same task. They are not.

Kansas asks 25 questions and requires 20 correct, so the allowance is 5. Wisconsin asks 50 and requires 40, so the allowance is 10. Identical percentage, double the room. That is the first thing the subtraction reveals and the percentage hides.

Now push it one step further, into the part that actually changes how each of them should prepare. On the Kansas paper, one wrong answer costs 4 percent of the score. On the Wisconsin paper, one wrong answer costs 2. Meaning the Kansas applicant loses a fifth of their entire margin with a single miss, and the Wisconsin applicant loses a tenth.

Think about what that does to a weak topic. Suppose both of them are unreliable on parking and turning rules, and suppose those rules make up roughly a tenth of each paper. That is around two or three questions in Kansas and around five in Wisconsin. The Wisconsin applicant can get every single one of them wrong and still pass, because five is inside an allowance of ten. The Kansas applicant getting all of theirs wrong has spent half the margin on one topic, and any two further slips anywhere on the paper end the attempt.

So the two of them should not study the same way. Wisconsin's applicant is best served by broad coverage, because no single topic can sink them and the risk is spread thinly across a long paper. Kansas's applicant needs to find and close their worst topic specifically, because on a short paper with a small allowance, one unclosed gap is the failure mode.

The takeaway: the percentage tells you the bar, the allowance tells you the risk, and only the allowance tells you what to do next.

SourcesKansas Department of Revenue, Kansas Driving Handbook, 20 of 25 correct, 80 percent (checked 2026-08-27); Wisconsin Motorists' Handbook (BDS126), Wisconsin DMV, 40 of 50 correct, 80 percent (checked 2026-08-27)

Published Margins and Rhode Island's Estimate

Set the state figures side by side and the range is wider than most people expect. Published margins run from three spare answers to ten, while Rhode Island's working figure would reach twelve. The percentage is a poor predictor of where a state lands.

Look at Rhode Island's estimate and Maryland's published mark together. Rhode Island's DMV publishes 40 questions but no pass mark; the Community College of Rhode Island, which gives the DMV test as the final exam of its driver education course, tells students they need 70 percent, which would be 28 of 40 and 12 spare answers, the most generous margin in the set. Maryland requires 22 of 25, which is 88 percent and 3 spare, the tightest. On that working figure a Rhode Island applicant can miss four times as many questions as a Maryland applicant on a paper that is only fifteen questions longer.

Idaho is the case that shows percentage and allowance pulling apart in the middle of the range. It sets 34 of 40, an unusually high 85 percent, yet the length of the paper still leaves 6 spare, which is more room than Kansas gives at 80 percent. Massachusetts does the reverse from below, setting the lowest published bar in this group at 18 of 25, about 72 percent, and still leaving only 7.

None of these states is being generous or harsh on purpose. Each one picked a paper length and a bar independently, and the margin is whatever fell out of the pair. Which is exactly why the number has to be looked up rather than inferred.

• Rhode Island: 28 of 40 if CCRI's 70 percent applies, 12 spare; the DMV publishes no pass mark.
• Wisconsin: 40 of 50, 80 percent, 10 spare.
• Massachusetts: 18 of 25, about 72 percent, 7 spare.
• Idaho: 34 of 40, 85 percent, 6 spare.
• Kansas: 20 of 25, 80 percent, 5 spare.
• Pennsylvania: 15 of 18, about 83 percent, 3 spare.
• Maryland: 22 of 25, 88 percent, 3 spare.

SourcesRhode Island Division of Motor Vehicles, Rhode Island Driver's Manual, 40 multiple choice questions, 90 minutes; no pass mark stated (checked 2026-08-28); Community College of Rhode Island, Driver Education, 70 percent to pass the RI DMV Knowledge Test given as the course final (checked 2026-08-28); Massachusetts RMV, Learner's Permit Knowledge Test FAQs, 18 of 25 correct, about 72 percent (checked 2026-08-27); Idaho Transportation Department, Idaho Driver's Manual, 34 of 40 correct, 85 percent (checked 2026-08-27)

Break One: A Second Bar Inside the Paper

The first thing that invalidates a clean subtraction is a state that scores part of the paper separately. When that happens the overall allowance still exists, but it can no longer be spent anywhere you like.

New York asks 20 questions and requires 14, so the arithmetic says 6 spare. It also requires at least 2 of the 4 road sign identification questions correct. Miss three signs and the attempt is over even with 15 correct answers overall, because the sign bar is checked on its own.

Georgia splits the paper in two and scores each half. Its published mark is 15 of 20 on each section, 30 of 40 in total. Nominally that is 10 spare, but they are not pooled: five misses in one section fails that section regardless of a perfect score in the other. Indiana works the same way, requiring 80 percent or higher on each of two components rather than across the whole exam. Michigan states only that both sections must be passed and publishes no percentage at all.

The practical rule is that whenever a state describes its test as having parts, sections or components, stop treating the total as a single pot. Find the bar for each part, subtract inside each part, and take the smallest of the resulting margins as the one that governs you.

SourcesNew York DMV, 14 of 20 to pass, with at least 2 of the 4 road sign questions correct (checked 2026-08-27); Georgia Department of Driver Services, Test and Exams Information, 15 of 20 on each section, 75 percent, 30 of 40 overall (checked 2026-08-27); Indiana Bureau of Motor Vehicles, Knowledge Exam, 80 percent or higher on each of the two components (checked 2026-08-28); Michigan Department of State, New drivers (18 and older), both sections must be passed; the state publishes no percentage (checked 2026-08-27)

Break Two: Your Age or Your History Changes the Paper

The second thing that invalidates the subtraction is looking up the wrong test. Some states run more than one, and the version you sit depends on facts about you rather than about the state.

California is the clearest case. An applicant under 18 sits a 46 question paper requiring 38 correct, which leaves 8. An applicant of 18 or older sits a 36 question paper requiring 30, which leaves 6. Two people at the same counter on the same morning have different papers and different margins.

Utah splits on history rather than age. Its standard test is 50 questions requiring 40, but an applicant previously licensed elsewhere takes a 25 question open book test requiring 20. Open book does not mean easy, but it does mean the allowance of 5 is measuring something completely different from the allowance of 10 next to it.

Before you trust any figure you have found, check which of these buckets you are in. A number quoted without an age or a history qualifier attached to it is a number that may not be yours.

SourcesCalifornia DMV, 38 of 46 for applicants under 18; 30 of 36 for applicants 18 or older (checked 2026-08-27); Utah Driver License Division, Written Knowledge Test, 40 of 50 standard; 20 of 25 open book for applicants previously licensed elsewhere (checked 2026-08-27)

Break Three: One of the Two Numbers Is Not Published

The third break is the most awkward, because there is nothing you can do about it except know that it is happening. Some agencies publish a pass mark and no question count, and at least one publishes a question count and no pass mark.

Colorado's rule states 80 percent as the official pass mark and the state publishes no question count to apply it to. Mississippi's administrative code sets 80 percent to pass without a published question count either. New Mexico's driver procedures manual sets 70 percent or better and, again, no count. In each of those three, the percentage is real and the allowance is unknowable from official sources.

North Dakota runs the other way round: the Department of Transportation publishes no pass mark on its learner permit page at all. Any allowance quoted for a state in this group is somebody's estimate, including ours, and it should be labeled as one rather than presented as the state's rule.

If you are testing in one of these states, the honest plan is to stop hunting for a number that has not been published and prepare as though the margin is small. That is not defeatism, it is just what you do when the specification is incomplete.

SourcesColorado DMV Rule 8, 1 CCR 204-30, Class R Driver Testing and Education Program, 80 percent is the official pass mark; Colorado publishes no question count (checked 2026-08-27); Mississippi Administrative Code, Title 31 Part 1, Driver Services, Rule 2.3, 80 percent to pass; no question count published (checked 2026-08-28); New Mexico MVD Driver Procedures Manual, Chapter 11, Written Test Requirements, 70 percent or better to pass; MVD publishes no question count (checked 2026-08-28); NDDOT, How to Apply for a Learner's Permit, NDDOT publishes no pass mark (checked 2026-08-28)

Break Four: The State Publishes the Allowance Directly

The last one is a break in the pleasant direction. Montana's administrative rule does not state a pass mark to be converted. It states the answer you were trying to calculate: 7 or fewer wrong on the combined signs and rules test.

That phrasing is worth pausing on, because it is a better piece of drafting than a percentage. A percentage has to be turned into a count, and the turning is where rounding disputes and misquoted figures come from. A limit expressed in wrong answers cannot be misread, and it is stated in the same units the applicant is actually thinking in while they sit there.

If your own state publishes its rule this way, you are done. Take the number as written, do no arithmetic, and do not adjust it against a percentage from anywhere else.

SourcesAdministrative Rules of Montana 23.3.121, Road Signs and Road Rules, 7 or fewer wrong answers on the combined signs and rules test (checked 2026-08-28)

What to Do With the Number Once You Have It

Turn the allowance into a study decision rather than a comfort blanket. Sit a full practice paper under real conditions, mark it, and count two things: how many you got wrong, and how many you got right without being sure. Add those together. If the total is larger than your allowance, you are not ready yet, and the size of the gap tells you how much work is left.

That second count is the one people skip and the one that matters most on a small margin. A confident guess that happens to land correctly is a question you would lose on a different draw from the same bank, and on a three answer margin two of those are the whole thing.

Then look up what a failed attempt actually costs you where you live, before you need to know. What one application fee covers and how long you wait before retaking are both set state by state, and knowing them in advance turns a short attempt into a scheduling inconvenience instead of a surprise.

• Sit a full length paper without notes and without pausing.
• Count the wrong answers and the unsure right answers as one total.
• Compare that total with your allowance, not with a percentage.
• Work only the topics behind those questions, then re-sit.
• Check your state's retake wait and fee before test day, not after.

Frequently asked