How to score 99 percentile in VARC

An honest word about the percentile first

A percentile is a rank, not a score. It tells you how many candidates you beat, so it depends on the paper's difficulty, the size and strength of the cohort in your slot, and the normalisation across slots — none of which you control. That is why no honest tutor can guarantee you a 99. What you can control is your accuracy and your attempt discipline, and those are what put you in the conversation for the top band year after year.

So reframe the goal. Instead of "hit 99 percentile", aim to be the kind of reader whose RC accuracy is so high that a 99 is on the table if the paper cooperates. That is a target you can train toward every day, and the rest of this article is about becoming that reader.

Why RC accuracy dominates the VARC percentile

RC is where the marks are. With about 16 of 24 VARC questions coming from four passages, Reading Comprehension is roughly two-thirds of the section — a larger and, crucially, more improvable pool than Verbal Ability. Para-jumbles, odd-one-out, and summary questions are fewer, and their accuracy is notoriously spiky even for strong candidates. RC rewards a trainable skill: reading an argument accurately and testing options against the text.

RC also compounds. Every hour you spend on RC accuracy sharpens the reading you bring to summary and para-jumble questions, because they are the same comprehension skill applied to a smaller text. The reverse is not true — drilling para-jumbles does little for your passages. If time is limited, RC is the highest-return place to spend it. The full sequencing sits in our CAT VARC preparation guide, but the headline is simple: get RC accuracy up, and the section score moves with it.

A realistic attempt strategy across four passages

You do not need to attempt all 16 RC questions to reach the top band. You need to attempt the ones you can get right. Most high scorers triage the four passages rather than marching through them in order. Here is a workable default you can adapt to your own accuracy data.

• Spend the first 60–90 seconds triaging. Skim the opening lines of all four passages and rank them by how readable they feel. Start with the two that look most tractable — you want early marks in the bank, not a brutal abstract-philosophy passage eating your first twelve minutes.
• Do two to three passages thoroughly. At roughly 8–10 minutes each, three full passages is most of your 40-minute VARC window. Read once for the argument, then answer with option elimination, committing to an answer on every question you attempt.
• Cherry-pick the fourth. On the hardest passage, take only its most answerable questions — usually a main-idea or a specific-detail question — and leave the ambiguous inference or tone item you would only be guessing on. A skipped question costs nothing; a wrong one costs you a full mark plus the negative.

The exact split — whether you attempt 12, 14, or 16 — should come from your own record, not a number you read online. If your data shows 90 percent accuracy at 13 attempts and 78 percent at 16, the 13-attempt plan scores higher after negatives. Work that out on your mocks, then trust it. For the mechanics of choosing per question, see how to improve RC accuracy.

Build an error log — it is the whole engine

The single habit that separates a stalled 92 percentile from a climbing 98 is a maintained error log. Every wrong RC answer is data. On its own it is noise, but thirty logged errors reveal the two or three patterns actually costing you the section. Then you stop practising RC in general and start attacking your specific leaks.

For each RC question you miss — and each one you guessed and happened to get right — record four things:

Review the log every weekend. If ten of your last thirty errors are inference questions where you over-reached, that is not a mystery — it is next week's assignment. A practice tool that explains every option, not just the key, makes the log almost effortless to fill, because the reason each distractor was wrong is already spelled out for you.

Daily reading plus daily testing

RC accuracy is built by two habits running in parallel: reading widely to become comfortable with dense, unfamiliar arguments, and testing yourself to convert that comfort into marks. Neither works alone. Reading without testing leaves you fluent but untested against traps; testing without reading leaves you brittle the moment a passage strays from your comfort genres.

Read one demanding essay a day from a rotation — The Economist, Aeon, The Atlantic, Scientific American, Nature, Harvard Business Review, Project Syndicate — deliberately choosing genres you find hard. Then do at least one timed RC passage and review every option. CAT passages are drawn from exactly these kinds of arguments, so the more abstract-philosophy and dense-economics prose you have metabolised, the fewer passages can ambush you. A concrete day-by-day version is laid out in our daily RC practice routine.

The bottleneck used to be supply: finite question banks get exhausted and half-memorised, which quietly inflates your accuracy and lies to you on test day. PracticeRC removes that ceiling — paste any article or essay and it generates a fresh, CAT-style RC test with options, instant scoring, and an explanation for every option. You never run out of unseen passages, so your accuracy numbers stay honest.

Avoid the negative-marking traps

CAT penalises wrong answers, and the top band is often lost not to the questions you failed but to the ones you should never have attempted. Three disciplines protect your score.

• Do not answer to feel productive. A blank costs nothing; a wrong answer costs a mark plus the negative. If you have eliminated only one option and are choosing between three on a coin-flip, that is a skip, not an attempt. Guessing is only rational once you are down to a genuine two-way split.
• Beware the "true in the real world" option. The most seductive distractor is a statement you know to be true from general knowledge but that the passage never supports. RC tests the text, not your knowledge. If you cannot point to the line, do not pick the option.
• Respect the sunk-cost trap. Six minutes into a passage you dislike, the instinct is to force answers to justify the time. Resist it. The minutes are already spent; do not add wrong marks to the loss. Take the answerable questions and leave.

For the anatomy of each distractor type and how to disqualify it on evidence, read RC question types explained. Naming the flaw in every rejected option is what turns "this feels off" into a defensible skip or a confident answer.

A phased plan to build it

You cannot do everything at once, and trying to is why most plans collapse in week three. Build the skill in phases, each with one job.

Throughout, keep the daily reading going in the background — it is the slow-compounding half of the work. You can run unlimited passages for every phase on CAT reading comprehension practice, and the full preparation arc, including Verbal Ability, is mapped in the CAT VARC preparation guide.

Frequently asked questions