modafinil (Provigil)

no boxed warning; the rash warning comes first

the first warning on the label is about the skin: "serious rash requiring hospitalization and discontinuation of treatment has been reported in association with the use of modafinil" [label 5.1]. the trial figure behind it is from children. in modafinil’s pediatric trials, rash that led to stopping occurred in about 0.8% (13 of 1,585), including one case of possible stevens-johnson syndrome and one apparent multi-organ hypersensitivity reaction, with a median time to the rash of 13 days, and "no such cases were observed among 380 pediatric patients who received placebo" [label 5.1]. after approval, rare serious or life-threatening rashes, including stevens-johnson syndrome, toxic epidermal necrolysis and dress, were reported in adults and children, and the label says the reporting rate of the two worst, itself "generally accepted to be an underestimate", "exceeds the background incidence rate" of 1 to 2 cases per million person-years [label 5.1].

the label then explains why its instruction is blunt. no factor predicts who will get a serious rash; nearly all cases came 1 to 5 weeks after starting, but some came after about 3 months; benign rashes also happen and "it is not possible to reliably predict which rashes will prove to be serious". so: "modafinil tablets should be discontinued at the first sign of rash, unless the rash is clearly not drug-related", with the caveat that stopping "may not prevent a rash from becoming life-threatening or permanently disabling or disfiguring" [label 5.1]. two related warnings: angioedema and anaphylaxis, seen in trials of armodafinil (its r-enantiomer) and reported after approval with modafinil [label 5.2], and multi-organ hypersensitivity reactions, "including at least one fatality in postmarketing experience", typically fever and rash with liver, heart or blood involvement, a median of 13 days after starting [label 5.3]. hypersensitivity to modafinil or armodafinil is the label’s only contraindication [label 4].

the other warnings, in plain words

beyond the rash warnings, the label carries:

• sleepiness may not go away: people with abnormal sleepiness "should be advised that their level of wakefulness may not return to normal", should be reassessed often, and may need to avoid driving; the label notes that patients "may not acknowledge sleepiness or drowsiness until directly questioned" [label 5.4].
• psychiatric symptoms: in adult trials, anxiety (1%), nervousness (1%), insomnia, confusion, agitation and depression (each under 1%) led to stopping more often than on placebo. after approval, reports included "mania, delusions, hallucinations, suicidal ideation, and aggression, some resulting in hospitalization", many but not all in people with a psychiatric history; one healthy volunteer developed paranoid delusions and hallucinations on 600 mg a day with sleep deprivation, gone 36 hours after stopping. the label says to use caution with a history of psychosis, depression or mania, and to consider stopping if psychiatric symptoms develop [label 5.5].
• heart: chest pain, palpitations, breathlessness and ecg changes occurred in three trial subjects with mitral valve prolapse or left ventricular hypertrophy, and one man with a history of fainting had a 9-second pause in heartbeat after 27 days at 300 mg a day. the label does not recommend modafinil in left ventricular hypertrophy or in mitral valve prolapse that reacted badly to stimulants before. average blood pressure did not change in trials, but more people on modafinil needed new or increased blood-pressure medicine, 2.4% against 0.7% (3.4% against 1.1% in sleep apnea) [label 5.7].
• driving: modafinil "has not been shown to produce functional impairment", but the label asks for caution with driving or machinery until a person is reasonably sure it does not affect them [label 5.6].
• hormonal birth control: modafinil speeds the breakdown of drugs cleared by the liver enzyme cyp3a4, and "the effectiveness of steroidal contraceptives may be reduced when used with modafinil tablets and for one month after discontinuation of therapy". the label recommends another or an additional method during treatment and for a month after [label 7]. it also lowers cyclosporine levels, and can raise levels of drugs cleared by cyp2c19, such as phenytoin, diazepam, propranolol, omeprazole and clomipramine, and, in people who lack the cyp2d6 enzyme, some tricyclics and ssris; it asks for closer inr checks with warfarin and caution with maois [label 7].
• pregnancy: the label reports intrauterine growth restriction and miscarriage with modafinil and armodafinil, says "whether the cases reported with modafinil are drug-related is unknown", and records developmental toxicity in rats and rabbits at exposures near human doses [label 8.1].
• older people and the liver: the label suggests lower doses and close monitoring in older people, whose elimination may be slower, and half the dose in severe liver impairment [label 2.3, 2.4, 8.5].

misuse and dependence

the label is plain that modafinil acts like a stimulant in some ways: it "produces psychoactive and euphoric effects, alterations in mood, perception, thinking, and feelings typical of other cns stimulants", binds the dopamine reuptake site and raises extracellular dopamine, and was self-administered by monkeys trained on cocaine. in a study in people experienced with drugs of abuse, 200 to 800 mg produced effects "consistent with other scheduled cns stimulants (methylphenidate)". prescribers are asked to watch closely anyone with a history of drug or stimulant misuse, for rising doses or drug-seeking [label 9.2]. on dependence, one trial watched for withdrawal for 14 days after 9 weeks of use and saw none, "although sleepiness returned in narcoleptic patients" [label 9.3]. in overdose, there have been postmarketing reports of fatal overdoses involving modafinil alone or with other drugs [label 10].

what the trials measured, in whom, and for how long

narcolepsy: two 9-week us trials randomised 558 adults to 200 mg, 400 mg or placebo. the main measure was the maintenance of wakefulness test, minutes to fall asleep while trying to stay awake in four 20-minute sessions, together with a clinician’s global rating. starting from about 6 minutes, people on modafinil stayed awake about 2.0 to 2.3 minutes longer by the end, while those on placebo fell asleep about 0.7 minutes sooner; 58% to 72% were rated improved, against 37% and 38% on placebo [label 14.1, tables 2, 3].

sleep apnea: in a 12-week trial of 327 adults who were still sleepy despite cpap, the wakefulness test rose by 1.5 to 1.6 minutes from about 13 minutes, against a 1.1-minute fall on placebo, and 61% to 68% were rated improved against 37%. in a 4-week trial of 157 adults, the epworth sleepiness scale fell 4.6 points on 400 mg against 2 on placebo, from about 14 [label 14.2, tables 2, 3].

shift work disorder: in a 12-week trial of 209 night-shift workers, time to fall asleep during a simulated night shift rose 1.7 minutes from 2.1 on 200 mg, against 0.3 minutes on placebo (p<0.05), and 74% were rated improved against 36% [label 14.3, tables 2, 3]. the label adds that "not all patients with a complaint of sleepiness who are also engaged in shift work meet the criteria" for the disorder [label 14.3].

what the numbers mean. every trial beat placebo, and the gains were real but bounded: in narcolepsy the wakefulness test moved from about 6 minutes to about 8 to 9 out of a possible 20, and the label warns that wakefulness "may not return to normal" [label 5.4, table 2]. 400 mg was "well tolerated" but there is "no consistent evidence that this dose confers additional benefit beyond that of the 200 mg/day dose" [label 2.1].

the uses it was never approved for, from its own label

the label reports two pediatric programmes that did not lead to approval. in a 6-week trial in 165 children and adolescents with narcolepsy, modafinil did no better than placebo on a sleep test or on a clinician’s rating, and side effects included tourette’s syndrome, insomnia, hostility, more cataplexy and hallucinations on falling asleep, and suicidal ideation [label 8.4]. in three 7- to 9-week adhd trials in children and adolescents aged 6 to 17, at up to 425 mg a day, modafinil did beat placebo on a school-version adhd rating, "there were 3 cases of serious rash including one case of possible sjs among 933 patients exposed", and the label closes: "modafinil is not approved for use in treating adhd" [label 8.4].

the label reports no trial of modafinil for depression, for fatigue in other illnesses, or for thinking or memory in people without a sleep disorder; none of those is an approved use [label 1, 14]. its only mention of depression trials is in the overdose section, where two people in "foreign depression studies" took 4,000 and 4,500 mg without life-threatening effects [label 10].

side effects, with the label’s numbers

in pooled placebo-controlled trials of 934 people on modafinil and 567 on placebo across all three disorders, the commonest were headache 34% against 23%, nausea 11% against 3%, nervousness 7% against 3%, rhinitis 7% against 6%, back pain 6% against 5%, diarrhea 6% against 5%, anxiety 5% against 1%, insomnia 5% against 1%, dizziness 5% against 4% and dyspepsia 5% against 4%; also loss of appetite 4% against 1%, chest pain and raised blood pressure each 3% against 1%, and depression, palpitations and fast heart rate each 2% against 1% [label table 1]. headache and anxiety rose with dose. 8% stopped because of side effects against 3% on placebo, most often for headache [label 6.1]. two liver enzymes, ggt and alkaline phosphatase, ran higher on modafinil, rarely outside the normal range [label 6.1]. after approval, agranulocytosis, a severe fall in white cells, was reported [label 6.2].

what the research says, in both directions

a limit to state first: this site’s library holds no row tagged modafinil yet. the research below is the label’s own trials and the one cross-drug review in the library that includes it.

in favour. all five approval trials beat placebo on their main measures, in three different disorders, and nighttime sleep (daytime sleep, for shift workers) measured in the sleep lab was unaffected [label 14]. the trial that looked for withdrawal after 9 weeks found none [label 9.3]. in the largest network meta-analysis of adhd medicines, 133 double-blind trials, modafinil was one of only two drugs (with methylphenidate) that beat placebo on teacher ratings in children and adolescents, a standardised mean difference of 0.76 [cortese-2018].

against, or smaller than it sounds. the same meta-analysis found that in adults modafinil did not beat placebo on adhd symptoms (standardised mean difference 0.16 in the wrong direction, 95% ci crossing zero), that adults were about four times as likely to stop it for side effects as placebo (odds ratio 4.01, 95% ci 1.42 to 11.33), and that amphetamines did better than modafinil head to head [cortese-2018]. in children the adhd benefit came with the serious-rash cases the label reports, which is the context for "not approved for use in treating adhd" [label 8.4]. the approval trials ran 4 to 12 weeks, the wakefulness gains were a few minutes on a 20-minute test, and wakefulness "may not return to normal" [label 5.4, 14]. the label’s abuse study found euphoric effects like methylphenidate’s [label 9.2].

questions people ask

does modafinil (provigil) have a black box warning?

no. the modafinil label has no boxed warning. its first warning is serious rash, including stevens-johnson syndrome: in pediatric trials 0.8% (13 of 1,585) had a rash that led to stopping, against none of 380 on placebo, and the label says to stop modafinil at the first sign of rash unless it is clearly not drug-related. it also warns about angioedema, multi-organ hypersensitivity with at least one death, psychiatric symptoms, persistent sleepiness and heart problems in people with certain heart conditions.

what is modafinil approved for?

improving wakefulness in adults with excessive sleepiness from narcolepsy, obstructive sleep apnea (alongside cpap, not instead of it) or shift work disorder. it is not approved for adhd, depression, fatigue in other conditions, cognitive enhancement, or for children. the label itself says it is not approved for adhd and not approved in children for any indication.

does modafinil affect birth control?

yes, according to its label. modafinil speeds the breakdown of drugs cleared by the liver enzyme cyp3a4, and the label says the effectiveness of steroidal contraceptives such as ethinyl estradiol may be reduced while taking modafinil and for one month after stopping. it recommends an alternative or additional method of contraception for that whole period.

is modafinil addictive or a controlled substance?

it is a schedule iv controlled substance. the label says it produces psychoactive and euphoric effects typical of cns stimulants, raises dopamine, and in a study in experienced drug users produced effects like methylphenidate’s, and asks prescribers to watch people with a history of stimulant misuse closely. one trial that watched for withdrawal for 14 days after 9 weeks of use found no withdrawal symptoms, although sleepiness returned in people with narcolepsy.

does modafinil work for adhd?

it is not approved for adhd. its label reports three trials in children and adolescents in which it beat placebo on a school adhd rating, but with 3 cases of serious rash, one possibly stevens-johnson syndrome, among 933 exposed, and states that modafinil is not approved for treating adhd. a 2018 network meta-analysis of 133 trials found it better than placebo on teacher ratings in children, but not better than placebo in adults, where people were about four times as likely to stop it for side effects.

what are the most common modafinil side effects?

in pooled placebo-controlled trials: headache 34% against 23% on placebo, nausea 11% against 3%, nervousness 7% against 3%, anxiety and insomnia each 5% against 1%, plus rhinitis, back pain, diarrhea, dizziness and indigestion. headache and anxiety rose with dose, and 8% stopped because of side effects against 3% on placebo.

sources for this section

the library rows link to our one-page summary of each study, with the original paper, its funding and our evidence grade on that page.

1. modafinil tablets prescribing information, fda label on dailymed (set id 944daf47-49a6-93ca-e348-4a0f1b6d937e, version 21, effective july 21, 2026; a generic of provigil): sections 1, 2.1 to 2.4, 4, 5.1 to 5.7, 6.1, 6.2, 7, 8.1, 8.4, 8.5, 9.1 to 9.3, 10, 14.1 to 14.3 and tables 1 to 3
2. drugs@fda, nda 020717 (provigil, modafinil), approved december 24, 1998
3. ahrq medical expenditure panel survey, prescribed medicines public-use files 2014 (hc-168a) to 2024 (hc-254a): modafinil has no reportable year (name masked in the public file for 2014 and 2021 to 2024; fewer than 60 respondents in 2015 to 2020); method on the usage section of this page
4. cortese s, et al. comparative efficacy and tolerability of medications for attention-deficit hyperactivity disorder in children, adolescents, and adults: a systematic review and network meta-analysis. lancet psychiatry 2018 (the library row is written up for methylphenidate; the modafinil findings were read from the pubmed abstract, pmid 30097390, on 2026-10-13) (library row)

Most-reported reactions

• Fatigue2,949
• Nausea2,248
• Drug ineffective2,140
• Headache1,870
• Fall1,794
• Depression1,619
• Anxiety1,503
• Somnolence1,419
• Dizziness1,404
• Pain1,291