How I Finally Broke 520 on the MCAT
I spent six weeks going through three different question banks before I realized I was doing it wrong. The problem wasn't the volume of questions. It was how I was treating them. Most people approach MCAT practice questions like they're flashcards — do a set, check the answer, move on. That's why their scores plateau around 505 and stay there. Here's the actual process that worked for me. First, take a full-length diagnostic test before you touch anything else. Not a practice question set. A full timed exam. AAMC is the only thing that actually matters for this — Kaplan, UWorld, Blueprint, whatever you want for content review, but your baseline needs to come from official material. My diagnostic score was 498. Roughly where I expected based on undergrad grades, but the breakdown told a different story. Chemistry/Physics: 12th percentile. CARS: 89th percentile. Biology/Biochemistry: 58th percentile. I knew I had a verbal strength, which meant I could lean into it strategically, but the C/P section was the real bottleneck. After the diagnostic, I stopped trying to build a perfect study schedule and started building a question review system. This is where most people fail. They'll do 40 questions and spend maybe ten minutes checking answers. That's not enough time. You should be spending roughly equal time reviewing as you are answering. For each question you get wrong, write down three things: what the question was actually testing (not what you thought it was testing), why you chose the wrong answer (process of elimination failure? content gap? misread?), and the principle that would have led you to the right answer.
I kept these in a simple spreadsheet. Row per question. Columns for topic, question type, error category, and the core concept. After three weeks of this, the spreadsheet had about 400 entries and I could already see patterns. The biggest one: I was consistently missing questions that asked me to compare two competing theories or models. The MCAT loves forcing you to pick between two partially correct statements. Once I identified that pattern, I adjusted my approach. I started reading every answer choice fully before eliminating, instead of jumping to reject choices as soon as I spotted something slightly off. The second pattern was a content gap in biochemistry kinetics that I hadn't realized I had. I could solve basic enzyme problems but whenever the question introduced an inhibitor type I hadn't memorized thoroughly — mixed inhibition, noncompetitive — I'd guess wrong. I spent two days just on inhibition types and came back to practice questions with a noticeable improvement. This is the counter-intuitive part that nobody talks about enough: doing more questions won't fix a specific content gap. You need to pause and close the gap directly before going back to question practice. I wasted about a week grinding questions while still missing the same inhibition problems repeatedly. Stopping to actually study the content for forty-eight hours fixed what dozens of question sets couldn't. For CARS, I had the opposite problem. My score was already high and it stayed high, but it also wouldn't go higher. The plateau at 89th percentile lasted four weeks. The workaround was realizing I was reading passages too carefully on the first pass and then second-guessing myself on answer choices that were technically supported but not the best answer. I started timing my initial read at under four minutes per passage and forced myself to commit to an answer on the first instinct for any question where two choices seemed plausible. Accuracy went down slightly during practice but my speed improved enough that I had extra time for harder questions, and the section score climbed to the 94th percentile by test day.
There are real limitations to this approach that you need to accept upfront. The spreadsheet method takes time — I'm talking about two to three hours of review for every forty-question block. If you're on a compressed timeline and only have six to eight weeks to study, you can't sustain that intensity across the entire prep period. In that case, prioritize quality over quantity. Do twenty questions, review them thoroughly, and move on. Better to master that process on a smaller scale than to burn out trying to do everything at full depth. Another limitation is that this method assumes you have access to a variety of question sources with different styles. AAMC questions have a particular flavor that no third-party bank replicates perfectly. If you're using primarily one non-AAMC resource, you'll develop good test-taking skills but you might find the actual exam's phrasing or answer choice structure jarring. I supplement my official AAMC materials with UWorld for content breadth and a few Blueprint sections for tougher questions, but I always make sure the majority of my practice comes from AAMC itself, especially in the final three weeks. The timeline that worked for me was roughly eight weeks. Week one was the diagnostic and building the review system. Weeks two through six were intensive question work with daily review sessions. I did between forty and eighty questions per day depending on how much content review time I needed. Weeks seven and eight shifted toward full-length practice exams — one every three days — with heavy review days between them. By the third full-length exam in week seven, my score stabilized around 518. The last full-length before the actual exam was a 521.
Get the Full Details
Test day itself was unremarkable in the way that means you didn't notice anything going wrong. The questions felt standard. Nothing unusually tricky on C/P that I hadn't seen before. CARS passages were dense but fair. I finished with about three minutes to spare on the last CARS passage, which was exactly the buffer I'd been training for. My final score was 520. The numbers on the score report tell you almost nothing about whether you're ready. What matters is whether your review process has caught the patterns in your mistakes. If you're making the same type of error repeatedly across different question sets, no amount of additional volume will fix it. You need to identify the specific mechanic of the mistake and adjust your approach to that mechanic directly. That's what separates people who break through from people who stall at the same score for months. I'm not saying this is the only way. Some people prep in four weeks. Some people prep for six months and never get above 510. But the question review system and the willingness to pause and close content gaps before continuing is what actually moved my score, and it's the part I see other students consistently skip over.
