A realistic 3-month GMAT study schedule, week by week.

“Study hard” is not a plan. It's a hope. Three months from now you'll either have a calendar you executed or a pile of half-finished practice sets and a test date you're dreading. The difference is almost never talent or even hours logged — it's whether the hours were sequenced. Fundamentals before drilling. Drilling before mocks. Mocks before the test. Each phase earns the next one.

This is a 12-week schedule built for the most common case: a working professional with roughly 10-12 hours a week, aiming for a meaningful score jump on the GMAT Focus Edition. It assumes weeknight sessions of 60-90 minutes and a longer block on the weekend for full sections and mocks. If your life looks different, the adaptation notes at the end show how to compress it to a month or stretch it to six.

The single biggest mistake in self-study is starting with content review and never measuring against the real exam until it's too late. You diagnose first so every hour after that is aimed at something specific.

The diagnostic-first principle

Before you open a single chapter, you need a baseline. Not a vague sense of “I'm okay at math, weak on reading” — an actual score and a per-topic breakdown. Without it, you'll spend week three reviewing arithmetic you already know and week ten discovering that Data Insights, the section nobody warns you about, is quietly wrecking your total.

The GMAT Focus Edition has three sections, each 45 minutes and each scored 60-90:

• Quantitative— 21 questions, Problem Solving only. Data Sufficiency is gone from this section.
• Verbal— 23 questions, Critical Reasoning and Reading Comprehension. Sentence Correction was removed.
• Data Insights— 20 questions: Data Sufficiency, Multi-Source Reasoning, Table Analysis, Graphics Interpretation, and Two-Part Analysis.

Total scores run 205 to 805. There is no essay and no Sentence Correction. You get a basic on-screen calculator on Data Insights only — none on Quant — and no formula sheet on test day. You can bookmark as many questions per section as you like, but you can change at most three of your answers in each section. Knowing the format cold is part of week one, because every minute you spend confused by the interface on test day is a minute stolen from the questions.

The ration that governs everything: six official mocks

There are exactly six official practice exams on mba.com, and they are the only trustworthy signal of your real score. Third-party mocks vary wildly in difficulty and scoring; some inflate, some deflate, and none of them adapt the way the real test does. Treat the six official exams as a non-renewable resource. This schedule uses them deliberately — one as a cold baseline, then spaced through the back half so each one measures a real change and feeds a week of repair.

Burn through all six in the first month “just to practice” and you'll arrive at the final stretch with no clean instrument left to tell you whether you're ready. Ration them.

Week 1 — Baseline and orientation

Take one official mba.com practice exam cold, start to finish, under real timing, before you study anything. It will feel uncomfortable and your score will probably be lower than you hoped. That's the point — this is the measurement, not the performance. A baseline taken after a week of “warming up” is a baseline you've already contaminated.

With the result in hand, do three things. First, learn the Focus format in detail: section order, question counts, the bookmark and answer-edit rules, where the calculator lives. Second, read your section and per-topic breakdown to see where the points actually leaked. Third, set a target score that is ambitious but real — usually 60-110 points above your baseline over three months — and write down the sub-targets per section that would get you there.

Weeks 2-4 — Fundamentals

This is the content phase. The goal is untimed competence: can you do the thing correctly when nobody's rushing you? Speed comes later and only after accuracy exists to be sped up.

Quant content

Rebuild the foundations in order: arithmetic and fractions, then number properties (factors, multiples, primes, divisibility, odd/even and positive/negative behavior), then core algebra (linear and quadratic equations, inequalities, exponents and roots). Don't memorize a wall of formulas — understand why each one works, because the test rewards recognizing structure, not reciting identities. Remember there is no calculator on Quant, so mental and paper arithmetic has to be fluent.

Verbal method

Verbal on the Focus Edition is Critical Reasoning and Reading Comprehension only. For CR, learn to dissect an argument into its parts: the conclusion, the evidence that supports it, and the unstated assumption bridging the two. Most CR question types are just different operations on that skeleton — strengthen, weaken, find the assumption, identify the flaw. For RC, build a reading method that captures structure over detail: what is each paragraph doing, what is the author's attitude, where does the passage turn. You answer from the structure, not from rereading.

Intro to all five Data Insights types

Spend real time here because Data Insights is where self-studiers under-invest most. Get a working introduction to each of the five: Data Sufficiency (a logic skill more than a math skill — you decide whether information is enough, not what the answer is), Multi-Source Reasoning, Table Analysis, Graphics Interpretation, and Two-Part Analysis. You won't master them in three weeks; you just need to stop being surprised by their formats.

Throughout these three weeks, do light daily question sets — ten to fifteen questions tied to whatever you studied that day — untimed or loosely timed. The point is repetition of the new method while it's fresh, not a timed gauntlet.

Weeks 5-8 — Deliberate practice

Now you convert competence into reliability. This is the longest phase because it does the most work. You drill topic by topic at increasing difficulty, and — this is the part most people skip — you start an error log and tag every single miss.

Topic-by-topic drilling, escalating difficulty

Take one topic at a time and work it from medium up to hard. When you can clear a topic at hard difficulty with steady accuracy, move on. Rotate across all three sections each week so nothing goes stale. Begin layering in timing now — not white-knuckle speed, but a soft sense of the per-question budget so it's familiar before the integration phase.

The error log, and why it beats raw volume

Every question you miss gets logged with a tag for whyyou missed it: a concept you don't know, a careless slip, a misread, a timing failure, a bad strategy choice. Tagging turns a vague pile of wrong answers into a ranked to-do list. If half your misses are careless, more content review won't help — a checking routine will. If they're conceptual, you have specific gaps to close.

Review beats volume. Doing a thousand questions and reviewing none of them teaches you almost nothing. Doing three hundred and dissecting every miss teaches you almost everything.

Spaced retrieval of past misses

Start cycling old misses back in. A question you got wrong two weeks ago should resurface to confirm the fix actually stuck. Spaced retrieval — revisiting material at widening intervals — is one of the most evidence-backed learning techniques there is, and it's the difference between understanding something once and owning it. Around week six, take one timed section mock (a single 45-minute section, not a full exam) to feel real timing under pressure without spending one of your six official full-length exams.

Weeks 9-11 — Integration

Topic competence and full-test performance are different skills. You now have the first; this phase builds the second. The enemies here are stamina, section switching, and pacing under real fatigue.

Mixed sets under timing

Stop practicing one topic at a time. Do mixed sets that jump between topics and difficulties the way the real test does, always on the clock. This trains the mental gear-shifting the exam demands — the thing that makes question 15 feel harder than it is because you're tired and it followed something unrelated.

A full official mock about every 10 days

Take a full-length official mock roughly every ten days through this stretch — figure two to three exams here. Simulate test conditions: same time of day you'll test, same section order, no pausing, real breaks only. The score matters, but the debrief matters more.

Repair the week after each mock

This is the engine of the back half. After each mock, spend the following days fixing exactly what the debrief exposed. If Data Insights timing collapsed, drill Data Insights pacing. If you missed a cluster of geometry, rebuild geometry. Don't take another mock until you've repaired the last one — an untreated weakness will just show up again and waste your next official exam.

Week 12 — Taper

Athletes taper before a race, and so should you. Take your final official mock early in the week — ideally five or six days out — so you have time to act on it without panic. Then ease off. Light review of your error log and a handful of confidence-building questions, not marathon sessions.

Do not cram. The night before the test, do nothing GMAT-related. You cannot meaningfully raise your score in the final 48 hours, but you can absolutely lower it by arriving exhausted and anxious. Sleep, logistics, and a calm morning are worth more than any last-minute drill.

The 12-week calendar at a glance

Compressing to one month

If you only have four weeks, the schedule doesn't get shorter so much as denser. Roughly double your weekly hours to 20-24 and collapse each phase: one week of baseline plus rapid fundamentals, two weeks of aggressive deliberate practice with the error log running from day one, and a final week of integration and taper. Cut your mock budget to two official exams — one early as a baseline, one late as a readiness check — and lean on section mocks in between. The non-negotiables stay non-negotiable: diagnose first, tag every miss, ration the official exams. A compressed plan that skips the error log isn't fast, it's just shallow.

Stretching to six months

With six months, halve the pace rather than doubling the content. Give yourself two full passes through fundamentals instead of one — the first to learn, the second to consolidate after a few weeks of drilling have shown you what you actually misunderstood. Extend the deliberate-practice phase so you reach hard difficulty in every topic with margin to spare, and let spaced retrieval run over a longer horizon, which is where it does its best work. You still have only six official mocks, so spread the back-half exams to roughly every two to three weeks, each followed by a full repair cycle. More time should buy depth and retention, not more passes over material you've already mastered.

The short version

Diagnose in week one with a cold official mock. Build fundamentals in weeks two through four. Drill deliberately with an error log in weeks five through eight. Integrate under timing with full official mocks in weeks nine through eleven, repairing after each. Taper in week twelve and don't cram. Ration your six official exams, because they're the only honest score signal you have. Review beats volume, and a tagged miss you revisit is worth ten you forget.

The platform

This is the concrete calendar that complements the study-plan framework — the framework explains the why, this gives you the dated what. Zakarian GMAT runs the whole loop for you: an official practice-exam baseline you enter on the platform, chapters that build fundamentals, topic-filtered practice with escalating difficulty, a six-tag error log, a spaced-retrieval review queue, and full-length mocks scored on the 205-805 Focus scale with a mock-to-mock trend. The sample chapter is free if you want to see the teaching first.