Danish Citizenship Test Pass Rate 2016–2026: Who Passes and Who Fails

Roughly one candidate in two fails the Danish citizenship test. Across the ten sittings since the harder 45-question test began in November 2021, the official pass rate has averaged 49%, swinging between 35.8% and 58.6%. This guide sets out every official figure since 2016, the counts behind them, the pattern that summer sittings have usually gone better than winter ones, and the ministry’s one detailed breakdown of who passes — including candidates from India, Pakistan and other countries.

UIM Integrationsbarometer (source SIRI); SIRI news, 19 August 2026.

The mean of the ten post-reform pass rates is 49.25% — our calculation from the barometer figures. The number sitting the test has roughly doubled over the same period, from about 3,200 in June 2022 to 6,242 in June 2026, so the absolute number passing each year has risen even though the rate has not.

Since the reform, the summer sitting has had the higher pass rate in three of the four years with both: 58.0% against 35.8% in 2023, 52.1% against 41.4% in 2024, and 58.2% against 47.2% in 2025. The exception was 2022 (47.3% in June, 53.0% in November). The ministry’s November 2023 analysis links that winter’s record low to a larger share of refugees, of candidates under 25 and of repeat takers — people who had failed in May and tried again six months later. There is no official statement that the November papers are harder. The difference is more plausibly about who sits in November than about the paper itself.

A pass needs both 36 correct answers and 4 of the 5 values questions. The statistics show that it is the 36-correct rule that stops most people. In June 2026, 91% of candidates met the values rule but only 59% reached 36 correct; in November 2025 the figures were 79.5% and 48.6%. Of those who failed in November 2023, about 62% missed only the count rule, 6% only the values rule and 33% both (rounded). In SIRI’s random samples, questions were answered correctly 72.8–80% of the time on average (78.3% in June 2026), just under the 80% line.

The ministry published one detailed analysis of who passes, for the November 2023 sitting, comparing it with May 2023. It is the only official source that splits results by nationality, and no later sitting has been broken down the same way — so treat it as a picture of one difficult sitting, not a permanent ranking.

The clearest single signal is Danish level. Among candidates who had come to Denmark as adults, those holding Prøve i Dansk 3 passed at 64%, against 21% for holders of Prøve i Dansk 2 and 15% for Prøve i Dansk 1. The test is written at PD3 level, so reading speed and vocabulary matter as much as knowledge. Candidates who had failed in May and tried again in November passed at only 19%, against 41% for those who had not sat the May test — six months of the same preparation rarely changes the outcome; a different preparation does.

A few cautions before drawing conclusions from these numbers. The pass rate is the share of those who actually sat the test, not of those who registered — in June 2026, 6,593 were registered according to SIRI’s analysis annex and 6,242 sat it, so a few hundred people did not turn up. The counts for 2022 and May 2023 are derived by dividing the number who passed by the pass rate, because SIRI published only the number passing for those sittings. The May 2023 figure differs slightly between sources: SIRI’s own news says 58.0% (2,605 passed), while the Integrationsbarometer series gives 58.2%; we use SIRI’s figure in the table and the barometer’s in the chart. And the barometer labels the June 2026 sitting “May 2026”, although the test was held on 3 June. None of these differences changes the picture, but they explain why two websites quoting “official” figures can disagree by a decimal point.

The November 2023 breakdown also needs care. Its groups overlap — an Indian citizen who came to Denmark for work, holds PD2 and is sitting the test for the second time appears in four different rows — so the figures cannot be added up or compared as if they were independent. They show which characteristics went with lower pass rates in one sitting; they do not predict an individual result. What they do show consistently is that preparation and language level matter more than origin.

Three practical conclusions follow from the official data. First, the risk is real: across ten sittings, about half of candidates failed, and a fail costs the fee and six months. Second, the risk is concentrated in reaching 36 of 45 overall — the 35 book questions plus the 5 current-affairs ones — not in the values block, so time spent on the book, the past papers and the news pays off most. Third, if your Danish is at PD2 rather than PD3 level, build reading time into your preparation — the ministry’s figures show that language level predicts the result more strongly than nationality does. Our guides on the learning material and the past papers show where the 35 questions come from.