Loading…
The dMAT isn't a knowledge test with a syllabus to memorise. It has a fixed set of task typesto get fluent with. Here's exactly what the exam looks like for applicants in India — every module, every subtest, the timings, the scoring and the dates.
That distinction matters more than it sounds. Most exams an Indian graduate has sat reward coverage: there is a body of content, and the work is getting through it. The dMAT has almost no content at all. What it has is four task formats, a clock that is genuinely tight, and a working environment more restrictive than the one you practise in. A candidate who has never seen a Latin square finds the first one close to impossible and the tenth one mechanical, and the difference between those two moments is exposure rather than ability.
This page is the map of what you are preparing for: the structure of both modules, what each subtest actually asks, how the time is divided, how the result is reported, and which parts of the process sit outside the exam room. Everything here is drawn from the official information published by g.a.s.t. and APS India, and every date, fee and timing is theirs to change — confirm the details for your own sitting on the official sources at the foot of the page before you plan around them.
The whole format fits on one screen. Everything below this table is an expansion of a line in it.
The exam has a Core module measuring general study aptitude through three subtests, and a Subject module for applied reasoning. Indian and general applicants take the General Academic Subject module. Each module runs about 90 minutes (the Core module bundles its three ~25-minute subtests with on-screen instructions), so the whole exam is roughly three hours of testing plus a break — all single-choice, in English, delivered on a computer.
g.a.s.t. states that the examination lasts three and a half hours in total including that break, which is the figure worth planning your day around rather than the sum of the two module timings. The two halves ask for different things from you. The Core module wants you to hold and manipulate information mentally across three unrelated task types. The General Academic module wants you to read carefully and reason from a text. Switching between those demands is not automatic, which is why the second module is where avoidable marks tend to go — by then you have already spent your freshest attention.
The break itself is worth a plan. It is the one stretch of test day when you are not answering questions, and most candidates arrive at it with nothing in mind, sit down, and spend it re-running the module they have just finished. That review is worthless — nothing about a completed module can change — and it consumes exactly the attention the next ninety minutes needs. Having something specific to do in the break is the simplest way to stop that spiral starting. Rehearse it in practice rather than improvising it on the day.
The Core module measures how you reason under time pressure, not what you have memorised. Its three subtests look nothing alike, and in your head they are not alike either: Figure Sequences is visual and holistic, Mathematical Equations is symbolic and sequential, Latin Squares is spatial and constraint-driven. Each has its own tempo and its own way of loading your working memory. What they share is that all three reward the same broad habits — identifying structure, applying a rule accurately, and avoiding unnecessary work.
Spot the rule transforming a matrix of abstract figures, then pick the figure that completes it.
Solve small systems of linear equations in your head — the letters stand for whole numbers.
Fill a 5×5 grid so every row and column holds each letter once, using constraint logic.
Figure Sequences feel unpredictable on first contact because the information is visual rather than verbal. There is nothing to recall and nothing to calculate — the task is to reduce a busy image to a small number of testable changes, then apply the same change once more. Strong performance comes from checking a fixed set of properties in a consistent order, not from waiting for the answer to announce itself.
Drill this format on its own with free Figure Sequences practice questions.
This subtest does not test advanced mathematics. It tests whether you can reason cleanly with simple relationships while a clock runs, which is a different skill and a more trainable one. The fastest solve almost always comes from noticing a relationship between the equations rather than grinding through any one of them.
Drill this format on its own with free Mathematical Equations practice questions.
The rule is simple to state and the difficulty is entirely in applying several constraints at once without losing your place. A systematic process beats repeatedly rescanning the grid, and the systematic process is short enough to become automatic well before test day.
Drill this format on its own with free Latin Squares practice questions.
Most preparation happens one subtest at a time, which leaves the least practised part of the Core module to test day: the changeover. The screen moves to a different kind of task while your head is still in the previous mode, and two separate things go wrong. The first is carry-over — coming out of equations, candidates attack the first figure matrix analytically, decomposing every element in order instead of letting the pattern surface, and spend ninety seconds on an item they would normally read in twenty. The second is residue: you are still arguing with the block you have just finished, which from the inside feels like concentration.
The on-screen instruction between subtests is your seam, and a short fixed routine fits inside it — change your posture for one breath, name the mode you are entering in three or four words, and decide your first concrete move on item one. Anything longer gets abandoned under pressure. A finished subtest is finished: you cannot edit it, you cannot know how it scored, and your sense of how it went is unreliable in both directions.
Around 90 minutes of passage-basedquestions: you read a short academic text and answer single-choice questions that apply reasoning across mathematics, natural sciences, engineering, economics and social science. It tests transfer and application, not recall — there's no specialist knowledge to revise.
The official description emphasises academic problem solving, transfer and application, which is worth reading literally. The surface topic of a passage may be a field you have never studied; the skill being measured is using the information in front of you. A passage that introduces a technical term has to give you what you need to follow it, or its own questions would be unanswerable. The useful habit is to treat an unknown word as a label rather than a gap — call it “the process”, hold on to what the text says it produces and what it depends on, and carry on.
Across the module, a question may ask you to do any of the following:
Some questions hang on a chart or a table, and these reward a skill that is easy to underrate: reading quantitative information carefully instead of glancing at it. Read the frame before the data — the title, both axis labels, the units, the legend. A surprising share of avoidable errors come from confusing thousands with millions, percentages with counts, or one category's line with another's. Note whether an axis starts at zero, whether the figures are absolute or relative, and whether a total is already a percentage or something you still have to divide.
Then let the question tell you what to extract. For a single value, trace from the axis rather than estimating from position. For a difference, read both numbers exactly before subtracting; judging a gap by eye is where confident mistakes live. For a trend, describe the shape in plain words before you compare options. A quick estimate often rules out two or three options at once and saves the arithmetic entirely.
Single-choice items are written carefully, and the words that decide them are usually small ones: most, least, only, except, best supported, the passage implies rather than the passage states. One of those can make two options sound right when only one is. Read the stem, read it again before you look at the options, and register what it actually asks for. Where an option feels right but the stem feels off, the stem is usually the one telling the truth — and an option that uses words from the passage while changing the relationship between them is the most common trap in the module.
The timings are the part of the exam pattern that changes how you should prepare. Around 20 tasks in about 25 minutes leaves you a little over a minute per Core item — not five. Almost every candidate could answer most of these questions correctly given unlimited time, and the exam does not offer unlimited time, so a substantial part of what it measures is how you behave when the clock is genuinely tight.
Work out your budget before test day — section time divided by the number of tasks — not so you can hit it on every item, but so you develop an internal sense of when one is running long. Then decide your skip-and-return rule in advance: if a question clearly exceeds the budget, mark it, take the best option you currently have, and move on. Nothing punishes a score more than three minutes spent on one hard item and five easy ones never reached. Pacing is not about going fast everywhere; it is about going efficiently where you can, so there is room to think calmly where you cannot.
Length is its own difficulty. Three and a half hours is longer than most people ever practise for, and the final subtest is answered by a more tired person than the first. Difficulty that only appears in the third hour is invisible in twenty-minute practice sessions, and it is one honest reason a real score can land below what practice suggested.
The dMAT is delivered on a computer at a licensed g.a.s.t. test centre, and the Core subtests are designed to be solved mentally rather than on paper. You cannot bring your own notes or paper into the room, so confirm what writing material — if any — your centre provides, on the official sources and in your participant portal, and then practise inside whatever that turns out to be.
This is worth reading as a description of a working method rather than merely a rule. On paper you offload constantly without noticing: a half-formed rule goes in the margin, two eliminated options get a line through them, a partial sum sits at the top of the page. Take the page away and every one of those small deposits has to be held somewhere else. That is genuinely demanding, it has little to do with how good your reasoning is, and it is trainable on its own terms.
The mechanical losses are quieter but just as real. You cannot annotate a figure to mark what changed between two frames, or put a diagram beside its answer options. Your eyes travel between a figure and the options below it instead of taking in a whole page at a glance. The fix is structure rather than effort: scan in a fixed order so you always know where you have already been, use the cursor as a pointer, and name things instead of marking them — a rule you can say survives in your head far better than a shape you are trying to picture. The practical consequence for preparation is simple. Stop printing. Do your ordinary drills on a real computer, full-screen, one question at a time, without highlighting the file.
The certificate reports results in more than one way, and the two numbers answer different questions. The dMAT score converts your correct answers onto a scale running from 0 to 200, reported separately for the Core module and for the relevant subject module — the General Academic Module on the APS India route. The official description says the total results from adding the two module values together. The percentile rank compares your result with the other participants: on the official explanation, a percentile rank of 70 means 70% of participants achieved a lower or equally good result while 30% achieved a higher one. It does not mean you answered 70% of the questions correctly.
Standardising the scores is what makes the two halves comparable at all. Raw counts of correct answers from two differently built modules would not be worth comparing; figures placed on a common scale are. It is also why no honest preparation platform can promise you an official dMAT score from an unofficial mock. A practice platform can report raw accuracy, time, section performance and its own internal percentiles once it has enough data — it cannot reproduce the official standardisation process without official data, and anything presented as an official score prediction should be treated with suspicion.
Because the total is arrived at by addition, it hides the shape behind it. Two candidates can reach the same total by opposite routes, and the split between your two module scores is the only genuinely diagnostic thing the certificate offers. Treat a clear difference as information and a narrow one as noise. A strong Core with a weaker subject module usually points at coverage and familiarity rather than at how you think; the reverse — a respectable subject module with a weaker Core — is more often a question of pacing and method than of knowledge. If the two sit close together, the percentile rank is the more informative number.
Universities decide individually how they use dMAT results, so a programme may look at the overall result, one module, a percentile threshold, a ranking contribution, or a rule of its own. The only reliable source for how your result will be read is the admissions page of the specific programme you are applying to, for the current cycle. For the APS India route, the dMAT result is included as an additional element in the APS documentation, and APS India states that a low result does not automatically cause refusal of the APS certificate. The certificate itself does not expire: it is downloaded from the participant portal when results are released, and the original, unaltered file is what you submit.
For the first cycle, registration opened on 29 June 2026 and closed on 15 September 2026, with the sitting on 26 September 2026 and certificates available from 12 October 2026. The exam fee is €150. The official dMAT pages list ten test centres in India — Ahmedabad, Bengaluru, Bhopal, Chandigarh, Chennai, Kolkata, Mananthavady, Mumbai, New Delhi and Pune — with Kathmandu in Nepal and Berlin in Germany also available. Which of them still has a seat when you book is a separate question, so the list shown to you during registration is the one that counts. Dates, fees and centres are set by g.a.s.t. and can change; always confirm on the official sources below before you plan around them.
Two practical notes follow from this calendar. The first is that the deadline is a backstop rather than a plan: the steps that depend on other people — centre capacity, payment processing, document checks — are the ones worth doing early, because the centres available to you are simply the ones still open when you register. The second is that the choice of centre quietly sets your travel, your sleep and what condition you are in when the first module begins. A centre ninety minutes away on a route you have used before can genuinely beat one half that distance across traffic you cannot forecast. Exam-day information — the reporting time, the address, the meeting point and any centre-specific instructions — is normally published to your participant portal a few days before the test, and those instructions take precedence over anything you read elsewhere, including here.
The dMAT is an aptitude assessment administered by g.a.s.t.; APS India verifies academic documents. They are connected but not the same thing, and the dMAT does not replace APS verification — for affected applicants the certificate becomes one document inside the APS process. On the information currently published, the requirement applies to Master's applicants for the Summer Semester 2027 intake and later whose previous degree falls within Engineering; Commerce, Accounting, Finance or Economics; or Business or Management.
Transitional exemptions sit outside that requirement, and the headline one relates to APS registrations completed before 29 June 2026. They are short enough to skim and easy to misread, because each turns on a load-bearing qualifier — not a registration started but one completed. Our am I affected? check runs your field, intake and registration history against the current rules in one pass, and the official APS India pages settle anything it cannot.
Interdisciplinary and mixed degrees — Business Analytics, Industrial Engineering, Engineering Management and the rest of that family — sit on the boundary and are settled by the official affected-fields list rather than by a course title. One asymmetry is worth stating plainly: preparing for a test you turn out not to have needed costs a fee and some Saturdays, while believing you were exempt when you were not can cost an intake. Two other things follow from how the processes interlock. APS India has confirmed that affected applicants may submit their APS document package before the dMAT certificate is available, so the two workstreams should run in parallel rather than in sequence — but APS can only complete the procedure once the required certificate has been submitted and the result checked against g.a.s.t.
Half the value of a syllabus page is telling you what you can stop worrying about. For the dMAT, that list is long:
One consequence catches people out. Generic Indian aptitude books — RS Aggarwal, bank-exam material, CAT DILR sets — train a workflow built on rough paper and on working backwards from multiple-choice options, and the Core module gives you neither. They are not simply easier or harder than the dMAT; they rehearse a different way of working, which is why volume from them can leave your score exactly where it started.
Because there is nothing to cover, preparation is about fluency with four formats and the ability to hold them in working memory under a clock. Start from the official g.a.s.t. materials — the model tasks, the solution approaches and the interface videos — because they are the clearest available record of how instructions are phrased, how answers are selected and what a valid task looks like. Hide the solution before you attempt an example; reading it first creates false confidence, because the pattern is obvious once it has been shown.
Then build fluency on original practice, move it onto a screen, and add the clock last. Our eight-week prep guide sets out that progression week by week, the free practice questions let you drill one task type at a time, and a full timed mock tells you which of the four difficulties is actually yours: method, working memory, pacing or stamina. Those have different fixes, and there is no point applying one to another.
g.a.s.t. states the examination lasts three and a half hours in total, with a break between the two modules — roughly 90 minutes of testing each side of it. Confirm the exact schedule for your own sitting in the official information and your participant portal.
The Core module is made up of three subtests — Figure Sequences, Mathematical Equations and Latin Squares — each running to around 20 tasks in about 25 minutes. With on-screen instructions between them, the module occupies roughly 90 minutes in total.
There is no syllabus in the usual sense. The dMAT is an aptitude test rather than a knowledge test, so instead of topics to revise it has a fixed set of task types to get fluent with: three Core subtests and a passage-based General Academic Subject module.
Indian and general applicants take the General Academic Module, which runs for around 90 minutes. It is passage-based: you read a short academic text and answer single-choice questions that apply reasoning across mathematics, natural sciences, engineering, economics and social science.
Prepare as though you cannot. g.a.s.t. does not publish, on its public pages, what may be brought to the desk or what the centre provides — that is in the exam rules and participant instructions for your sitting, so confirm it there. The Core subtests are built to be answered without written working either way, which is why training your working memory matters more here than on most exams.
Each module is reported on a scale running from 0 to 200, alongside a percentile rank, rather than as a simple pass or fail. The official description says the total results from adding the two module values together. Your certificate does not expire.
It places you against the other participants rather than against a fixed bar. On the official explanation, a percentile rank of 70 means 70% of participants achieved a lower or equally good result and 30% achieved a higher one. It does not mean you answered 70% of the questions correctly.
Yes to both. The dMAT is delivered digitally at a licensed g.a.s.t. test centre, and the instructions, questions and answer options are all in English. Every task is single-choice.
For the first India cycle, registration opened on 29 June 2026 and closed on 15 September 2026, the test is on 26 September 2026, and certificates become available on 12 October 2026. Dates are set by g.a.s.t. and can change — confirm them on the official sources below.
The exam fee is €150. Fees are set by g.a.s.t. and can change, so check the official sources before you budget for it.
The requirement applies to Master's applicants for the Summer Semester 2027 intake onward whose previous degree falls in an affected field — Engineering; Commerce, Accounting, Finance or Economics; or Business or Management — unless a transitional exemption applies. Interdisciplinary degrees sit on the boundary and are settled by the official affected-fields list, not by the course title.
No. APS India states that a low dMAT result does not automatically cause refusal of the APS certificate. The result is an additional element in the APS documentation, and universities decide individually how they use it in admissions.
The syllabus above is the shape of the test. These are how each piece actually behaves once you are working through it.
Which of the three subtests to open first, and what to do when one of them stalls.
Read the rule before you look at the options, in a fixed order that stops you guessing.
The elimination order that gets you to the marked cell without filling the whole grid.
Substitution and elimination are usually slower than the shortcut the format allows.
What the passages ask for, and why reading faster is not the same as answering faster.
The questions are about what is stated, which is a narrower job than understanding the chart.
How the two module scores are reported, and what a percentile rank does and does not say.
We link out so you can verify everything yourself. We are not affiliated with any of these bodies.