Are Night Owls Smarter? What the Chronotype Studies Show

No. Night owls are not reliably smarter than early birds. Large studies find a small link between later sleep timing and cognitive test scores, and the best-known study behind the headlines used a vocabulary test and self-reported bedtimes rather than IQ and a body-clock measure. Whether you are an owl or a lark matters less for an IQ test than when in the day you sit it.

Where does the "night owls are smarter" claim come from?

Most coverage traces back to Satoshi Kanazawa and Kaja Perina, who published "Why night owls are more intelligent" in Personality and Individual Differences in 2009. They used Add Health, a US study that interviewed 20,745 adolescents in 1994 to 1995 and followed 15,197 of them in 2001 to 2002, when they were 18 to 28. Intelligence came from the Peabody Picture Vocabulary Test, converted to an IQ metric, taken in adolescence. Sleep came from four self-reported times: wake-up and bedtime on days with a fixed schedule and on days without one.

Two details are rarely repeated. The direction tested was adolescent vocabulary score predicting adult bedtime, not late bedtime predicting a higher score. And about one in five bedtime answers (20.2% on weeknights) were given as noon or half past noon, which the authors treated as a midnight mix-up and recoded. The paper also admits the vocabulary test is "a valid measure of verbal intelligence, not general intelligence", and that it has no direct chronotype measure. For why vocabulary behaves like a proxy for general ability, see our piece on vocabulary size and IQ.

How big is the night-owl gap in that study?

By the paper's weeknight figures, average bedtime rose with childhood IQ band:

• Very dull (under 75): 23:41
• Dull (75 to 90): 00:03
• Normal (90 to 110): 00:10
• Bright (110 to 125): 00:21
• Very bright (over 125): 00:29

That is 48 minutes between the top and bottom bands, and the two top bands did not differ significantly from each other on weeknights. After controls for age, sex, race, education, earnings, religion and hours worked, each IQ point was linked to 0.008 hours of later weeknight bedtime. Multiplied by 15, that is about 7 minutes per standard deviation of IQ, our arithmetic from the paper's coefficient. A standardised coefficient of .04 means IQ accounts for roughly 0.2% of the variation in bedtime, and the effect on weekend wake-up time was not significant. The authors describe the effect size and the explained variance as "very small".

Their explanation, the Savanna-IQ interaction hypothesis, says more intelligent people adopt evolutionarily novel habits, and they offer the lack of nocturnal activity in an encyclopedia of traditional cultures as evidence that night life is novel. That is an interpretation of the pattern, not something the data test.

What do larger and newer studies show?

An earlier study by Roberts and Kyllonen (1999), of 420 US Air Force recruits, also found evening types more likely to have higher intelligence scores. Later authors note that the group was skewed toward morningness, since recruits had spent six weeks on a uniform sleep schedule, and that its tests were given at different times of day. We read only its abstract.

The most useful summary is the 2011 meta-analysis by Preckel and colleagues, whose numbers we take from the same authors' 2013 paper. Eveningness correlated r = .08 with cognitive ability and morningness r = -.04, both small. For academic achievement the signs flip: morningness r = .16 across 13 studies and eveningness r = -.14 across 6. An r of .08 explains under 1% of the variance, so the ability edge for owls is tiny.

The biggest recent dataset is UK Biobank. West and colleagues at Imperial College London analysed 26,820 people aged 53 to 86 in 2024. On a global score from tests of reasoning, pairs matching, reaction time and prospective memory, evening types scored 0.135 standard deviations above morning types in one cohort (10,067 people) and 0.075 in the other (16,753), and intermediate types were close behind (0.106 and 0.063). On the usual scale that is roughly 2 and 1 IQ points, our arithmetic and not the authors', and these are mixed cognitive tests rather than an IQ test. The gap is as much about morning types scoring lowest as about owls scoring highest. The study is cross-sectional, uses one self-report question for chronotype and did not adjust for education, so it cannot say that being an owl causes better scores.

Smaller 2025 and 2026 papers point both ways, and we read only their abstracts. A 2026 study of 75 young adults in Experimental Brain Research found early chronotypes had shorter reaction times and fewer errors, especially in the morning. A 2025 study of 666 people found evening types reported more everyday memory lapses.

Why do night owls often get lower grades?

Grades reward showing up on time and other habits as well as ability. In a 2013 study of 272 German school students, eveningness predicted lower grade-point averages even after controlling for cognitive ability, conscientiousness and motivation, and chronotype was unrelated to cognitive ability in that sample. A 2014 study of undergraduates at Semmelweis University in Hungary (grades available for 247 of 753) found social jetlag, the gap between school-day and days-off sleep timing, lowered weekly average grades during term but not in the exam period, when students set their own hours. And when Seattle delayed its high-school start by nearly an hour, students slept a median 34 minutes longer and median grades rose 4.5% (Dunster and colleagues, 2018). That study covers all teenagers, not only owls, but it shows how much of a grade gap can be schedule rather than ability.

Does the time of day change an IQ-test score?

It can, mostly when your sitting clashes with your body clock. The synchrony effect is the idea that people perform best at their preferred time of day. In a 2023 study, Jankowski and colleagues tested 198 final-year students at 8:00 or around 13:00. Morning types beat evening types on a crystallized-intelligence test in the morning class (R² = .11, Cohen's d = .94, comparing the extreme groups) and not in the afternoon (d = .11). They saw no significant effect for fluid intelligence. An earlier adolescent study, by Goldstein and colleagues in 2007, found the opposite split: a synchrony effect on fluid measures, and a composite of Block Design, Digit Span and Vocabulary that worked out as a 6-point difference in IQ estimates between matched and mismatched times.

The two studies disagree on which ability is affected, and both are small, but the practical message is the same: sit a timed test when you are alert. Our pieces on sleep loss and test performance and on caffeine and test performance show other conditions that move a score, and our guide to what affects IQ test results and the page on sleep and IQ cover the rest. Repeat sittings matter too, as we explain in the practice effect on cognitive tests.

What does the claim get wrong?

• It says night owls have higher IQs, but the famous study used a vocabulary test, and its authors say it measures verbal, not general, intelligence
• It treats late bedtime as the cause, but the study predicted adult bedtime from adolescent score, and the UK Biobank work is cross-sectional
• It implies a big gap, but the effect sizes are very small: r = .08, 7 minutes per 15 IQ points, 0.075 to 0.135 SD
• It overlooks grades, where larks have the edge (r = .16 versus r = -.14)
• It forgets test time: evening types did worse on an 8:00 test in one study, which says nothing about their ability at their best hour

A fair summary is that chronotype and cognition are weakly related, in a direction that depends on the measure. If you are an owl, it does not predict a higher IQ, and if you are a lark, it does not predict a lower one. For comparison with other popular claims, read our fact-checks on the breakfast question and on zodiac signs and IQ.

What should you do before an IQ test?

• Pick a time you are normally alert, and avoid sitting a timed test at your sleepiest hour
• Sleep your usual amount the night before rather than trying to shift your body clock
• Treat any score as a range, not a point, using the margin of error in your report
• Do not read a single result as a verdict on whether you are a morning or evening thinker

To see how your own result sits in the population, take the IQ Metrics test, then check it on the IQ bell curve or the IQ percentile calculator.

Common questions