Three espressos' worth of caffeine didn't sharpen self-control, not even in people with more ADHD symptoms

Caffeine is the world's most widely used mind-altering substance, and most people take it to think better. For people with attention-deficit/hyperactivity disorder (ADHD) there is a more specific idea: caffeine acts a little like the stimulant medicines used to treat the condition, so it might help them more than anyone else. A large survey of US soldiers found that those with adult ADHD used more caffeine pills, energy drinks and caffeinated drinks, and that heavier use went with fewer of some symptoms. But surveys can't tell whether caffeine helps, or whether people who find it pleasant simply drink more of it. A trial published online on 4 October 2026 in Human Psychopharmacology: Clinical and Experimental put the idea to a direct test.

The researcher, from the University of Groningen in the Netherlands and Pamukkale University in Türkiye, recruited 36 students aged 18 to 30, 19 men and 17 women. Nobody was picked or turned away because of a diagnosis; instead each person filled in the 18-item Adult ADHD Self-Report Scale, and ten of them scored above the screening line for probable ADHD. Everyone came to the lab twice, about three weeks apart. Once they drank 200 ml of decaf coffee with 250 mg of pure caffeine stirred in, and once the same coffee with 250 mg of lactose. Three people split the jobs so that neither the participant, the person running the tests nor the person analysing the data knew which cup was which. The plan, outcomes and analysis were registered in advance.

Forty-five minutes after the drink, participants did two classic self-control tasks. In a go/no-go task they pressed a key for every letter except X, which came up about a third of the time and meant stopping a ready response. In a flanker task they reported which way a central arrow pointed while the arrows around it sometimes pointed the other way. Caffeine made no difference to either. Participants pressed on an X on 14.3% of X trials after placebo and 15.0% after caffeine, and their average response time was 396 versus 402 milliseconds. Flanker speed and accuracy were the same too. ADHD symptom scores didn't change the picture: across all 15 pre-planned tests of whether caffeine helped more as symptoms rose, none came close, and Bayesian statistics gave moderate support for no effect rather than a missed one. The tasks weren't simply too blunt: people with higher symptom scores tended to do worse on them, and a poorer night's sleep went with slower responses.

The study has clear limits. It was small, enough to rule out a moderate benefit but not a tiny one. It tested a single fairly large dose in rested young people who mostly drank little coffee (about 77 mg of caffeine a day on average), and most of them admitted having had some caffeine in the 24 hours before a session, though leaving those people out didn't change the result. Participants weren't asked how alert they felt, so there is no direct check that they noticed the dose. One unplanned measure hinted at slightly easier conflict handling on the flanker task, which the author treats as a lead, not a finding. And it says nothing about regular moderate coffee drinking, or about adults with diagnosed ADHD. What it does show is that, for students with a range of ADHD traits, one strong coffee is not a quick fix for impulse control.