Do Smart Drugs and Nootropics Raise IQ? What the Trials Measured

No. No smart drug or nootropic has been shown to raise IQ in healthy people, and the studies that exist did not measure IQ. What the best meta-analysis of modafinil, methylphenidate (the drug in Ritalin) and amphetamine did measure was how well rested, healthy adults performed on tasks such as recall, sustained attention and inhibitory control after a single dose, and the effects it found were small: standardised mean differences of 0.12 to 0.21. This guide sets out what the trials measured, puts the effect sizes on the IQ scale as an analogy, covers the one place IQ tests have been used (children with ADHD), and checks the over-the-counter brain supplements. It is an evidence review, not medical advice.

What does the best evidence on modafinil, methylphenidate and amphetamine show?

Roberts, Jones, Sumnall, Gage and Montgomery pooled 47 placebo-controlled studies of single doses in healthy adults who were not sleep-deprived (European Neuropsychopharmacology, 2020; the search ended in November 2019). They ran one meta-analysis per drug and report standardised mean differences (SMD), the gap between drug and placebo measured in standard deviations.

• Modafinil: 14 studies, 64 effect sizes, overall SMD 0.12 (95% CI 0.02 to 0.21); the only single domain with an effect was updating of working memory, SMD 0.28 (0.02 to 0.54) from 5 studies
• Methylphenidate: 24 studies, 47 effect sizes, overall SMD 0.21 (0.09 to 0.32), driven by recall (0.43, 0.20 to 0.65), sustained attention (0.42) and inhibitory control (0.27, 0.02 to 0.51)
• d-Amphetamine: 10 studies, 27 effect sizes, overall SMD 0.21 (-0.06 to 0.47), not statistically significant

The authors' verdict is blunt: there is a user perception that these drugs are effective cognitive enhancers, but this is not supported by the evidence so far. They describe the effects as small, probably transient, and measured in experiments that do not reflect how the drugs are used in the wider population. Two cautions cut the other way. A null result is not proof of zero: the review found no statistical equivalence in most domains, and amphetamine's point estimate equals methylphenidate's with a wider interval. And 22 published papers could not be included because they had no extractable data, which the authors note.

Earlier reviews point the same way. A 2015 meta-analysis of 48 studies and 1,409 participants (Ilieva, Hook and Farah, Journal of Cognitive Neuroscience) found small effects of methylphenidate and amphetamine on inhibitory control and short-term memory and a medium effect on delayed memory, qualified by publication bias, and concluded that the effect on healthy cognition is probably modest overall. The authors add that healthy users may be taking stimulants for energy and motivation more than for cognition. Repantis and colleagues (2010) found that expectations of these drugs exceed their actual effects, with better memory on methylphenidate and better attention in well-rested people on modafinil. Battleday and Brem's 2015 review of modafinil found that most basic tests showed better executive function, only half showed gains in attention or learning and memory, and a few reported poorer divergent creative thinking, although more complex assessments looked more consistent.

What would those effect sizes be on the IQ scale?

An effect size is measured in standard deviations, and the IQ scale has a standard deviation of 15, so an effect of 0.12 would be 1.8 points, 0.21 would be 3.2 and methylphenidate's best domain, recall at 0.43, would be 6.5. That conversion is an analogy. A gain on a recall task is not a gain in IQ points, and no IQ test was used. It is useful for scale. A full-scale IQ score has a standard error of measurement of around 2 to 3 points (see our guide to the IQ score margin of error), so a true 1.8-point gain would be hard to tell from noise in one person. Sitting the same cognitive test twice raises the score by around a quarter of a standard deviation, 3.75 IQ-scale points, from practice alone, as our guide to practice effects explains. And the gains do not spread: in the same review methylphenidate showed no effect on task switching, updating, spatial working memory or selective attention.

Has any trial used an IQ test?

Not in healthy adults, as far as we could find. A search of Europe PMC abstracts for modafinil, methylphenidate or amphetamine in healthy people with Raven's matrices, fluid intelligence, IQ or a Wechsler test as the outcome returned nothing. That is not proof that none exists, because we did not search full texts. IQ appears in two other ways. In a retrospective analysis of two studies of 89 healthy students, Randall, Shneerson and File (2005) found that modafinil's benefits showed up in the group with a lower IQ (mean 106) and not in the group with a higher IQ (mean 115.5), and concluded that a high IQ may limit the detection of modafinil's positive effects. And in children with ADHD, where the drugs are used as medicines, IQ tests have been a measured outcome.

Zhang, Jin and Zhang (Journal of Clinical Psychopharmacology, 2011) gave the Wechsler Intelligence Scale for Children-Revised to 237 Chinese children with ADHD, 159 treated with methylphenidate and 78 untreated, at the start and six months later. Performance and full-scale IQ were higher in the treated group at the end (p below .05), but the study was not randomised, the abstract gives no IQ-point figures, and the authors wrote that the result should not be read as an increase in intelligence. A controlled but unblinded study of 67 medication-naive boys aged 6 to 13 (Hellwig-Brida and colleagues, 2011) retested the WISC-IV after 8 to 12 weeks and found that both groups improved, which is what a practice effect looks like. For how ADHD relates to IQ scores more broadly, see our guides to ADHD and IQ and high IQ and ADHD.

How common is non-prescribed use, and what is the legal status?

Interest is high. A pooled analysis of mostly US college students (Benson and colleagues, 2015) estimated that 17% (95% CI 13% to 23%) had misused stimulant medication, mainly for academic reasons and mostly obtained from peers with prescriptions; the authors note that studies lack a standard definition of misuse. A 2020 systematic review of 111 studies found self-reported rates in population samples ranging from 2.1% to 58.7%, and little evidence that non-medical use improves academic performance in people without ADHD. In the United States, amphetamine and methylphenidate are Schedule II controlled substances and modafinil is Schedule IV (21 CFR 1308.12 and 1308.14), so all three are prescription medicines with side effects and legal limits, not supplements.

Do over-the-counter brain supplements raise IQ?

The evidence is no stronger. In the GEM trial, 3,069 volunteers aged 75 or older took 120 mg of ginkgo biloba extract twice a day or a placebo for a median of 6.1 years; 277 people on ginkgo and 246 on placebo developed dementia (hazard ratio 1.12, 95% CI 0.94 to 1.33), and a separate GEM analysis found that ginkgo did not result in less cognitive decline. A 2018 Cochrane review of 28 studies with more than 83,000 participants found no evidence that any vitamin or mineral supplementation strategy for cognitively healthy adults in mid or late life has a meaningful effect; the only positive signals came from long-term antioxidant vitamins. One trial complicates the story: COSMOS-Mind (2023), in 2,262 older adults, found a small benefit of a multivitamin on global cognition (z = 0.07, 95% CI 0.02 to 0.12) and none for cocoa extract, using telephone-based tests rather than an IQ test.

Single ingredients have their own literature. See our guides to creatine, caffeine, iodine, brain-training apps and dual n-back training, which tested similar claims against similar standards.

What should readers take from this?

• Treat claims that a pill raises IQ as unproven: the trials measured recall, attention and inhibition, and we found none in healthy adults that used an IQ test
• Keep the size in mind: effects of 0.12 to 0.21 standard deviations are about 2 to 3 points on an IQ-style scale, within the noise of a single IQ test
• Retesting alone adds about a quarter of a standard deviation, so a before-and-after score on a drug proves little
• If sleep, food or exercise is the gap, those are better studied: see our guides to sleep loss and test performance, sleep and IQ and nutrition and IQ
• Prescription stimulants are medicines: discuss them with a clinician, not a website

Our IQ Metrics test reports on the standard scale, the IQ percentile calculator turns a score into a percentile, and our guide to whether you can improve your IQ covers what the evidence supports. To place a score against the population, use the IQ bell curve.

Common questions