How to Improve Your IMAT Score by 20 Points: Where the Marks Are

Article 3(4) of the 2026 decree gives the marks: 1.5 points for each right answer, minus 0.4 for each wrong answer, and 0 for each question left blank. The maximum is 90 points. Article 2(2) sets the paper: 60 questions, each with five options and one correct answer, in five sections. Article 2(4) gives 100 minutes for the whole paper.

Two facts follow from this table. First, biology and chemistry together are 38 questions, 57 of the 90 points: they have the most questions in which points can be lost, but where you personally can gain most depends on your own wrong answers and blanks, so set your priorities from your own sheet. Second, reading and logic together are only 9 questions. Even if all 9 went from wrong to right, the gain would be 9 × 1.9 = 17.1 points, so these two sections alone cannot give you 20 points. The same order also matters on ranking day. If two candidates have the same total, article 3(6)(a) ranks first the one with more points in biology, then chemistry, then physics and mathematics, then logic, then reading.

There are three ways to change one question, and each is worth a different number of points.

A blank turned into a right answer gains 1.5 points (from 0 to +1.5). To gain 20 points this way you need 20 ÷ 1.5 = 13.3 questions, so 14 questions: 13 would give only 19.5, and 14 give 21.0.

A wrong answer turned into a right one gains 1.9 points (from −0.4 to +1.5). To gain 20 points this way you need 20 ÷ 1.9 = 10.5 questions, so 11 questions: 10 would give 19.0, and 11 give 20.9.

A wrong answer turned into a blank gains 0.4 points (from −0.4 to 0). To gain 20 points only this way you would need 20 ÷ 0.4 = 50 fewer wrong answers. That is possible in arithmetic, but only on a paper with at least 50 wrong answers out of 60, all turned into blanks. For most papers, fewer wrong answers is a help rather than the whole 20 points.

So if you gain only by turning wrong answers or blanks into right answers, reaching at least +20 takes between 11 new right answers (all from wrong answers) and 14 (all from blanks). Including wrong-to-blank changes can change how many answers you need to change; work out your own mix. In this guide +20 is an illustrative goal: how fast any student improves depends on that student, and no official source gives a typical rate. Your own 2026 answer sheet shows which questions to work on.

Universitaly shows each candidate their own answer sheet and score after the test. From it you can count right, wrong and blank answers in each section. The three sheets below are our illustrations, not real candidates: they show how the same goal of +20 needs a different plan for different papers. The section sizes match the 2026 decree, and every total is computed from the marking.

This student answers nearly every question, including many blind guesses: 30 right, 24 wrong and 6 blank, which is 30 × 1.5 − 24 × 0.4 = 35.4 points. Illustration only.

This student answers carefully but leaves a long tail of blanks at the end: 28 right, 6 wrong and 26 blank, which is 28 × 1.5 − 6 × 0.4 = 39.6 points. Illustration only.

This student sits a solid paper with a mix of wrong answers and blanks: 38 right, 12 wrong and 10 blank, which is 38 × 1.5 − 12 × 0.4 = 52.2 points. Illustration only.

Sheet A loses most of its points to wrong answers: 24 of them cost 45.6 points compared with right answers. Its 20 points come from 11 wrong answers turned right, mostly in biology and chemistry, which means learning those topics properly rather than guessing. Sheet B loses most to blanks: 26 of them, many at the end of the paper. Its 20 points come from 14 blanks turned right, which is mostly a pacing problem plus some missing topics. Sheet C needs a little of both: 12 changes spread across four sections.

In all three sheets, most of the 20 points come from biology and chemistry. That is a choice we made for the examples: we put most of the wrong answers and blanks in the two sections with the most questions (38 of the 60). Your own sheet may look different, and your plan should follow your own errors, section by section. To do the same count on your own paper, enter your right and wrong answers per section in the free score calculator, or compare your score with a university’s past published list in the score check.

Each question has five options and one right answer. A uniformly random guess among five options has a 20% chance of being correct. Our arithmetic on article 3(4): the average value of that guess is 1/5 × 1.5 − 4/5 × 0.4 = 0.30 − 0.32 = −0.02 points. On average a blind guess is worth slightly less than a blank. Once you can honestly rule out one option, the average turns positive, and it grows with each option you rule out. The table assumes the remaining options are equally likely to be right; treat its last column as a suggestion for getting the most expected points, not a rule of the test.

The break-even point is a 21.05% chance of being right (0.4 ÷ 1.9). Above it, answering gives more expected points than a blank; below it, a blank gives more. Two warnings. First, these are averages: on a single question you either gain 1.5 or lose 0.4. Second, "ruling out" an option only counts if you are really sure it is wrong. If your eliminations are often wrong, your real chance is lower than the table says. Look at your 2026 sheet: in each section, divide your right answers by the questions you answered. If that number was low, practise eliminating options before you change how often you answer.

Guessing also moves your score much less than knowledge does. Ten guesses with four options open add only about 0.75 points on average; ten wrong answers turned right add 19. Use the guessing rule to stop wasting points, but put your study time into the topics.

The date of IMAT 2027 has not been published. The 2026 test was on 29 September 2026 (DM 1005/2026, art. 2(3)), so this outline assumes a test around the end of September 2027 and should be moved once the official date is out. It is our suggestion, not an official schedule. The order follows the points: biology and chemistry first, because they hold most of the marks, then physics and mathematics, then logic and reading. Plan the topics from Allegato A of the decree, the official syllabus, not from a list made by anyone else.

If 20 points is your goal and you have about eleven months, that is about 1.8 points a month (20 ÷ 11), or roughly one question a month answered right that you used to miss. That is only an average for planning, not a promise of progress. Measure progress section by section, not only by the total. A total can stay the same while one section goes up and another goes down, and the section numbers tell you where the next month’s study should go.

The three reasons in your error log point to three different fixes. "Did not know it" is a syllabus gap: put the topic into next month’s plan. "Knew it but slipped" is a checking habit: reread the question stem and the units before you mark an answer. "Ran out of time" is pacing: in 2026 there were 100 minutes for 60 questions, so about 1 minute 40 seconds each on average. A question that takes three minutes uses 80 seconds more than that average, time that then has to come from other questions.

A 2026 result cannot be carried into 2027: article 3(7) of the decree says being eligible refers only to the selection in progress and gives no right to admission in later years. So 2026 eligibility cannot be carried forward, and anyone applying for 2027 must follow the 2027 admission rules once they are published. Your 2026 answer sheet is the best study guide you own, because it shows exactly where your points went. Our retake guide walks through reading it section by section and setting a target from past published university lists.