CAT 2026 Exam Pattern and Syllabus: The Non-MCQ Trap

A Mistake This Guide Exists to Prevent

On one of my first SIMCATs I skipped every Non-MCQ question I wasn't fully sure about, on the assumption that a wrong Non-MCQ cost me the same -1 as a wrong MCQ. It doesn't - there's no negative marking on Non-MCQs at all, so an educated guess is free money I was leaving on the table. I only caught this because I went back and read the marking scheme properly instead of going on what I'd absorbed secondhand from other aspirants. That's the whole reason this is a reference page and not a strategy essay: the pattern is a fixed set of facts, and getting one of them wrong costs real marks before you've answered a single question.

Section-Wise Structure

CAT is a computer-based test split into three sections, each timed independently. You cannot move between sections or borrow time from one to another - this single rule shapes almost every strategic decision aspirants make, including how I now allocate my own mock time (see my in-exam time management writeup for how badly it goes when you don't respect it).

VARC - Verbal Ability & Reading Comprehension

• Questions: ~24, roughly 70% Reading Comprehension and 30% Verbal Ability (para-summary, odd-sentence-out, para-jumbles)
• Time: 40 minutes
• Format: Mix of MCQ and Non-MCQ (Type-In-The-Answer)

DILR - Data Interpretation & Logical Reasoning

• Questions: ~20, arranged into 4-5 sets
• Time: 40 minutes
• Format: Mostly MCQ, with 2-3 Non-MCQ sets per attempt

QA - Quantitative Aptitude

• Questions: ~22, spanning Arithmetic, Algebra, Geometry, Number System, and Modern Math
• Time: 40 minutes
• Format: Roughly 70% MCQ, 30% Non-MCQ

Quick-Reference Table

I keep this table specifically because the max-marks column is easy to misjudge: VARC's higher question count means it carries slightly more total marks on paper than DILR does, despite the two sections getting identical time. I didn't register that until I sat down and multiplied it out - it doesn't change how I attempt a section, but it means treating all three sections as "worth the same" by default is off by about 20% between the smallest and largest.

Marking Scheme

+3 for every correct MCQ or Non-MCQ answer. -1 for every incorrect MCQ answer. No negative marking on Non-MCQ (TITA) questions - which is exactly the rule I got wrong above. Once I started attempting every Non-MCQ I had even a rough elimination on, rather than only the ones I was fully certain of, my Quant score in particular moved: several of the Non-MCQ questions in Number System and Modern Math are set up so that a plausible range of answers is easy to narrow down even without solving cleanly.

Syllabus by Section

Quantitative Aptitude Syllabus

Arithmetic (Percentages, Profit & Loss, SI/CI, Time-Speed-Distance, Time & Work, Averages, Mixtures), Algebra (Linear & Quadratic Equations, Inequalities, Functions, Logarithms), Number System, Geometry & Mensuration, and Modern Math (Permutations & Combinations, Probability, Set Theory, Sequences & Series). This is also the section where my own mock data shows the clearest return on hours: my QUANT average rose from 26.8 across my first five logged mocks to 35.6 across my last five, the steadiest improvement of any section I track (full numbers in my mock log). Once these topics are conceptually solid, my Quant shortcuts guide covers the calculation-speed techniques that turn concept knowledge into exam-day speed.

DILR Syllabus

Data Interpretation (Tables, Bar/Line/Pie Charts, Caselets) and Logical Reasoning (Arrangements, Puzzles, Blood Relations, Direction Sense, Data Sufficiency, Games & Tournaments, Venn Diagrams). CAT rarely tests a "type" you can memorize in isolation - sets increasingly combine two or three of these in a single puzzle, which is a large part of why this is the one section where my own average barely moved despite deliberate practice. See my DILR practice strategy guide for what I'd do differently.

VARC Syllabus

Reading Comprehension (5-6 passages across humanities, business, science, and abstract philosophy), Para-Summary, Odd-Sentence-Out, Para-Jumbles. There is no fixed vocabulary list to memorize - CAT tests comprehension and inference, not word meanings in isolation. My own VARC scores across 15 mocks ranged from 3 to 39, which tells you this section rewards a calm, careful read on the day more than any fixed prep list - see the six RC question types I actually get tripped up by for specifics.

What Changes Year to Year (and What Doesn't)

The core structure - three sections, 40 minutes each, MCQ + Non-MCQ mix - has been stable since 2020. What shifts slightly year to year is the exact question count per section (it has ranged from 19-24 per section) and the DILR set difficulty. I've stopped reading rumor threads about "this year's pattern will be different" - the reliable move is preparing for the full syllabus above and adapting to whatever the real paper looks like on the day, which is exactly what timed mocks are for.

How I Actually Use This Page

I keep this pattern in mind every time I set up a sectional mock - 40 minutes is 40 minutes regardless of which topics show up, and knowing the question count going in changes how I pace the first read-through of a section. The planning question that comes after "I know the pattern" is time allocation, which is a separate problem - my month-by-month CAT 2026 study plan walks through exactly that.

Whatever your plan, the syllabus above is your checklist. I log my own sessions against these specific topics - not just "Quant" or "VARC" - because that's the only way I can see, months later, what's actually covered and what's still a gap. Sign up for Karma Yogi free to track it the same way.