TOPIK Reading Question Types: What 20 Past Papers Actually Test (Data Analysis)

This article is our own count, not a summary of someone else's guide. Below we show the method, the raw numbers, and what we think they mean for study order.

What we analyzed

Two caveats we should state plainly. First, our TOPIK I set covers every reading question in all 10 papers — the point total comes to exactly 1,000, which is 100 points × 10 papers. Second, our TOPIK II set is not complete: questions 42–43 are missing across every paper, and one paper is also missing 23–24. So read the TOPIK II percentages as "share of the 956 points we did cover," not as the official blueprint. Percentages for the types we do cover are unaffected by the gap, but the two missing questions are worth about 4 points per paper.

Question types were tagged from the actual prompt and answer choices of each question — for example, "무엇에 대한 이야기입니까?" (What is this about?) is a topic-match question, and "( )에 들어갈 가장 알맞은 것을 고르십시오" (Choose what best fits the blank) is a vocabulary or grammar gap-fill depending on what the choices are.

TOPIK I reading: where the 100 points actually are

The headline: detail comprehension plus vocabulary is 63.4 points out of 100. Everything else combined — main idea, grammar, ordering, inference, purpose — is under 37.

That has a blunt implication. If you're aiming for TOPIK I level 2 (140+ total across listening and reading), you do not need to be good at all nine types. You need to be reliable at two of them, and those two are the two that vocabulary size most directly drives. Detail comprehension questions are not reasoning puzzles at this level — they fail when you hit a word you don't know.

TOPIK I: what each question number tests

This is the part most study guides skip. Across our 10 papers, here's how stable each block was:

Read that table as a time budget. Questions 31–33 and 40–48 are 12 questions that never changed type in any of the 10 papers — that's the part you can drill to near-automatic. The mixed blocks (49–56, 59–70) are passage sets where two questions hang off one text, so the second question's type shifts depending on what the passage supports.

Practical read: if you're short on time before the exam, drill 31–33 (topic match) and 40–48 (detail + main idea) first. They are 100% predictable in form and they're worth roughly half the paper.

TOPIK II reading: a flatter distribution

TOPIK II spreads points much more evenly. The top two types are 36 of ~96 points, not 63 of 100. That's the real difficulty jump between the two exams — not just harder vocabulary, but more distinct skills you have to be competent at. You can pass TOPIK I by being strong at two things. TOPIK II punishes a narrow skill set.

Four types appear only in TOPIK II and catch people out:

• Text type (questions 5–8): identify what kind of writing this is — an ad, a notice, a product description. Short, fast, and easy points once you've seen twenty of them.
• Headline interpretation (25–27): you're given a newspaper headline like "한쪽으로 메는 가방, 허리 건강 '빨간 불'" (TOPIK II, 35th test, reading Q25) and asked which sentence explains it best. Headlines compress grammar and use metaphor ('빨간 불' = warning sign), so this tests a register you won't meet in textbook sentences.
• Sentence insertion (39–41): given a sentence, find the ㉠㉡㉢㉣ slot it belongs in. This is a discourse-connection skill, not a vocabulary one.
• Tone & attitude (23, 46, 50): how does the writer feel about this. Question 23 was this type in 9 of 9 papers where it appeared.

TOPIK II: what each question number tests

The back half (32–41, 47–48) is the most predictable stretch of TOPIK II reading and also where the passages get longest. Those are the questions to practice under time pressure, because running out of clock is what actually costs people points there.

What this means for how you study

Three conclusions we'd draw from our own numbers:

1. Vocabulary first, and it isn't close. In TOPIK I, vocabulary-in-context is 24.7 points directly — and detail comprehension, another 38.7 points, is mostly gated by whether you know the words in the passage. Grammar is 7.8. Starting with a grammar textbook when you have 500 words is optimizing the wrong 8%.

2. Practice by question number, not just by paper. Because types are stable per number, doing "questions 40–48 from ten different papers" teaches the pattern far faster than doing one whole paper start to finish. Same time spent, more repetitions of the same skill.

3. For TOPIK II, add breadth deliberately. Headline interpretation, sentence insertion and tone & attitude together are 12 points per paper, and none of them are skills you will accidentally acquire while grinding vocabulary. Budget separate practice for them.

We built our own app around the first conclusion. TOPIK Note orders its 7,000+ words by how often they actually appear in past papers rather than alphabetically — our count across those same papers found the top 1,000 words cover 98.6% of TOPIK I reading passages (full data here) — and includes 870+ reading questions from 21 past papers with English explanations, which is the dataset this article is built from.

FAQ

Written by the TOPIK Note team. We make one of the apps mentioned here, so treat the last section as what it is — but the counts above are from official past papers and you can verify any of them against a paper you own.