22 questions, 40 minutes, no syllabus. Improving here means improving at choosing, not solving.
DILR is 22 questions in 40 minutes, arranged in sets of roughly four to six questions each. You cannot attempt them all, and you are not meant to. The section is decided almost entirely by set selection — choosing the two or three sets you can actually crack and refusing the rest. Candidates who improve here usually improve at choosing, not at solving.
There is no syllabus. A set can be a table, a tournament, a seating arrangement, a scheduling puzzle, or something with no established name at all. This unsettles people who prepared by topic. The underlying skill is reading an unfamiliar structure, working out what governs it, and deciding quickly whether it will yield.
Spend the first two to three minutes scanning every set before attempting anything. You are looking for the ones where the constraints are clear and the data is manageable — not the ones that look most familiar. A set that looks easy and turns out to have a hidden branch will consume fifteen minutes and return nothing.
Most DILR errors are bookkeeping errors, not reasoning errors. Draw the grid, table or diagram before reasoning. Fill in what is certain, mark what is possible, and keep the two visually distinct. Trying to hold a partially-solved arrangement in your head is where careless mistakes come from.
Look for the most constrained element first — the person with only two possible slots, the value with one valid range. Constraint propagation resolves sets far faster than testing possibilities at random.
The single most expensive habit in DILR is continuing with a set because of the time already spent on it. Ten minutes invested does not make the eleventh minute more likely to pay. Set a hard internal limit, and when it passes, leave regardless of how close it feels.
Untimed DILR practice trains the wrong skill entirely. The clock is the difficulty.
Take four sets, spend three minutes choosing two, then solve. Review whether you chose well.
Unfamiliar formats are the point. Repeating one type builds false confidence.
Solve them afterwards untimed to learn whether leaving was correct.
Attempting two or three sets accurately usually beats attempting four sets partially. Accuracy carries disproportionate weight here because a half-solved set typically yields nothing at all.
Question counts and timing are in the CAT 2026 exam pattern; where DILR sits in the overall schedule is in the full CAT 2026 preparation plan.
Verified Real Ones who sat this exam recently. Zero commission from any college.
Find a Real One