The CAT 2026 VARC section has 24 questions in 40 minutes: 16 RC questions from 4 passages and 8 VA questions in TITA format. Across these 24 questions, there are exactly 8 distinct question types — 4 in the RC section and 4 in the VA section. Each type tests a specific skill and responds to a specific strategy. Aspirants who treat all 24 questions the same way leave marks on the table; aspirants who know exactly what each type is asking can switch approach in real time and score consistently.
This guide covers every format: what it tests, the approach that works, and the most common error aspirants make on each one.
Know exactly which of these 8 types you're losing marks on — your diagnostic report tells you.
Get My Score Estimate
The 8 question types at a glance
| Question Type |
Section |
Format |
Frequency |
Negative Marking |
| Main idea / central argument |
RC |
MCQ |
1 per passage (4 total) |
Yes (-1) |
| Inference and implication |
RC |
MCQ |
1-2 per passage |
Yes (-1) |
| Author's tone and attitude |
RC |
MCQ |
1 per passage |
Yes (-1) |
| Vocabulary in context |
RC |
MCQ |
1-2 per passage |
Yes (-1) |
| Para jumbles |
VA |
TITA |
3-4 questions |
No |
| Para completion |
VA |
TITA |
2-3 questions |
No |
| Odd-one-out |
VA |
TITA |
2 questions |
No |
| Para summary |
VA |
TITA |
1-2 questions |
No |
Pro Tip — TITA Advantage
All 8 VA questions have no negative marking. This changes your strategy: even a 40-60% confident answer on a VA question is worth attempting. On RC MCQs, only attempt when you can eliminate at least 2 wrong options, given the -1 penalty. Do not apply the same attempt threshold across both question types.
The 4 RC question types in detail
1. Main idea / central argument
What it tests
Whether you can identify the passage's primary thesis — the claim the entire passage is supporting, developing, or arguing against.
Strategy
Read the first paragraph fully and the last paragraph fully before answering this question. The main idea is almost always stated explicitly in one of these two locations. Eliminate options that are too narrow (cover only one section of the passage), too broad (extend beyond what the passage covers), or state a claim the passage only mentions as background.
2. Inference and implication
What it tests
Whether you can draw a conclusion one logical step beyond what the passage explicitly states. The correct answer is not stated in the passage — it is the logical consequence of what the passage does state.
Strategy
Find the passage sentence that the inference question references (usually specified via a quote or line number). Identify what the sentence is claiming. Then ask: what must be true if this claim is true? Correct inference options are conservative — they extend the passage claim minimally. Wrong options either repeat what the passage already said (not an inference) or extend far beyond what the passage supports (over-inference).
3. Author's tone and attitude
What it tests
Whether you can identify the author's emotional or intellectual stance toward the passage's topic, or toward a specific position mentioned in the passage.
Strategy
Track the author's language throughout — especially hedges ("purportedly / allegedly / claimed to be") that signal skepticism, and unhedged assertions ("clearly / in fact / the evidence shows") that signal endorsement. Most wrong tone options use words that are extreme versions of the correct tone: "dismissive" instead of "skeptical," "enthusiastic" instead of "appreciative." Match the intensity of the tone, not just the direction.
4. Vocabulary in context
What it tests
Whether you can identify the specific meaning of a word as used in the passage context — not its dictionary definition. All four options are typically valid meanings of the word in other contexts.
Strategy
Use the 3-strategy toolkit from the
vocabulary-in-context guide: surrounding sentence method (what is the author doing in this sentence?), contrast and concession signals (what must this word mean relative to its contrast?), and the substitution test (which option reads naturally when replacing the word?). Never answer from dictionary recall.
The 4 VA question types in detail
5. Para jumbles
What it tests
Whether you can arrange 4-5 jumbled sentences into a coherent paragraph by identifying logical sequence, pronoun references, and connector signals.
Strategy
Use the anchor sentence method and logic-chain algorithm from the
para jumbles advanced guide: identify the opener (no back-reference pronouns), identify the closer (consequence or summary signal), then build the middle using pronoun and demonstrative references. Do not use trial-and-error — it fails on 5-sentence sets. Target: 90 seconds per question.
6. Para completion
What it tests
Whether you can identify the sentence that logically closes a given paragraph — i.e., the sentence that best completes the paragraph's argument without introducing new ideas or contradicting its tone.
Strategy
Apply the 4-step elimination method from the
para completion guide: identify paragraph direction, eliminate options with new concepts, check tone match, verify clean closure. The 3 wrong-answer traps are: new concept, tone mismatch, and half-close (closes but then reopens with a qualifier). Target: 75-90 seconds per question.
7. Odd-one-out
What it tests
Whether you can identify which of 4-5 sentences does NOT logically belong in the same paragraph as the others. You type the number of the odd sentence.
Strategy
Read all sentences and identify the dominant theme — the topic that 3-4 of the sentences discuss. The odd sentence will either introduce a different topic, bring in a concept not set up by the other sentences, or break the logical flow of the group. Don't look for factual contradiction — the odd sentence may be factually related to the topic but logically misaligned with the specific argument the others are building. Target: 75 seconds per question.
8. Para summary
What it tests
Whether you can identify the best 1-2 sentence summary of a given paragraph — capturing the main claim without adding information, over-generalising, or focusing on a specific example rather than the central argument.
Strategy
Read the full paragraph and identify its main claim in one phrase before looking at options. Eliminate options that: (a) focus on examples or supporting evidence rather than the main claim, (b) introduce new ideas not in the paragraph, (c) over-generalise ("all X" or "always Y" when the paragraph makes a more limited claim), or (d) are too narrow (mention only one part of the paragraph's argument). The correct option is accurate and complete without being expanded. Target: 75 seconds per question.
Time allocation across all 8 types
Given 40 minutes for 24 questions, your per-question budget averages 100 seconds. But this average hides the important difference between RC and VA question difficulty and time requirements.
A workable allocation for the full VARC section:
- 8 VA questions: 10-12 minutes total (75-90 seconds each). No negative marking, so attempt all. VA questions come first in many aspirants' strategy because they're faster to complete and no-negative means zero risk on uncertain answers.
- 4 RC passages: 28-30 minutes total (7-8 minutes per passage). Each minute breaks down as roughly 3-4 minutes reading and 4-5 minutes for 4 questions. Use remaining time to review RC answers where you had 50-50 options.
For a minute-by-minute breakdown of how top scorers structure the full 40 minutes, the VARC 40-minute strategy guide covers passage ordering, VA-first vs RC-first sequences, and time recovery tactics when one passage runs long.
Use the Optima Learn practice questions bank to filter questions by each of these 8 types individually. Drilling one type at a time reveals your accuracy pattern more clearly than mixed-format mocks, and identifies the 1-2 types where targeted improvement has the highest mark impact.
Before Your Next Mock — Know These 8
RC types: main idea (first + last paragraph), inference (one logical step beyond the text), tone (hedge vs. unhedged language), vocabulary (context not dictionary).
VA types: para jumbles (anchor → chain, no trial-error), para completion (direction → new concept check → tone → closure), odd-one-out (which sentence breaks the dominant theme), para summary (main claim, not examples, no over-generalisation).
Know Your VARC Weakness Across All 8 Types
A diagnostic session with an Optima Learn mentor maps your accuracy by question type and builds a targeted improvement plan based on the types that cost you the most marks.
Book a Free CAT Strategy Call
Common questions answered
How many question types are there in CAT VARC?
CAT VARC has 8 distinct question types — 4 RC types (main idea, inference, author's tone, vocabulary in context) and 4 VA types (para jumbles, para completion, odd-one-out, para summary). The VARC section has 24 questions total: 16 RC questions from 4 passages (4 questions each) and 8 VA questions in TITA format.
Which CAT VARC question type appears most frequently?
Among RC types, inference and main idea questions appear most frequently — typically 1-2 of each per passage. Among VA types, para jumbles appear most frequently — typically 3-4 of the 8 VA questions are para jumble type. Odd-one-out and para completion each typically appear 2 times, and para summary 1-2 times.
Are CAT VA questions MCQ or TITA?
All 8 VA questions in CAT VARC are TITA (Type in the Answer) format — you type the correct sequence or select the odd sentence number. TITA questions have no negative marking. RC questions are MCQ with a negative marking of -1 for wrong answers. This means you should attempt all VA questions (even with partial confidence) but be more selective about RC questions where you're uncertain.
How should I order the 8 question types within the VARC section?
Most aspirants follow this sequence: attempt VA questions first (60-70 seconds each, no negative marking), then RC passages in order of difficulty (easiest passages first). Some aspirants prefer starting with RC to use peak concentration on the heavier cognitive load. The key is knowing your personal sequence before the exam — not deciding it during the exam.
Which VARC question type is hardest in CAT?
Para jumbles are typically rated hardest by aspirants who haven't practised them systematically — but they become highly solvable with the anchor sentence method. Among RC types, inference questions are hardest because they require drawing conclusions one step beyond what the passage explicitly states. Vocabulary-in-context is rated easiest by aspirants who know the 3-strategy toolkit, and hardest by those who try to answer from memory.
Each of the 8 question types has a deeper dedicated guide on Optima Learn. Start with the types where you lose the most marks based on your mock analysis. The score predictor shows how each 1-mark improvement in VARC affects your overall percentile — which helps you prioritise which question type to tackle first. For complete CAT 2026 preparation resources across all three sections, explore the Optima Learn preparation guide library.
