Modafinil side effects and interactions: what the labels say

Key points

• In the US label’s pooled trials (934 patients on modafinil, 567 on placebo), headache occurred in 34% vs 23% and nausea in 11% vs 3%; 8% vs 3% stopped for an adverse reaction [1]. The armodafinil label reports lower rates from separate trials [2].
• Both US labels warn of serious rash (Stevens-Johnson syndrome, toxic epidermal necrolysis, DRESS), angioedema, multi-organ hypersensitivity, psychiatric symptoms and cardiovascular events [1][2].
• Labels name interactions through liver enzymes: lower exposure to steroidal contraceptives and cyclosporine, higher exposure to CYP2C19 substrates, and closer INR checks with warfarin [1][3].
• Pregnancy studies conflict: major-malformation rates range from 2.0% to 13.1% across four studies, and the UK text says not to use modafinil in pregnancy [3][6][7][8][9].
• The labels rest on patients with sleep disorders. We found no long-term safety data in healthy people.

What is documented, and where

Modafinil (Provigil) and armodafinil (Nuvigil, its R-enantiomer; see modafinil vs armodafinil) are prescription medicines, so “research use only” does not apply. This article restates what regulators and safety studies record. Approvals, benefit evidence and legal status are in the modafinil guide and legal status by country.

As of 8 November 2026: in the US both labels cover excessive sleepiness in narcolepsy, obstructive sleep apnea and shift work disorder, and both drugs are Schedule IV [1][2]. The UK text (emc, revised 14 February 2024) covers adults with narcolepsy only [3], as does the EU after the Commission decision of 27 January 2011 [4]. We did not check armodafinil’s status in the UK, EU or Germany, the German product information, or national registers.

Most common adverse reactions

The US labels list reactions in pooled placebo-controlled trials in narcolepsy, sleep apnea and shift work disorder (section 6.1). The two trial sets differ, so the columns are not a head-to-head comparison.

Reaction	Modafinil (n=934)	Placebo (n=567)	Armodafinil (n=645)	Placebo (n=445)
Headache	34%	23%	17%	9%
Nausea	11%	3%	7%	3%
Nervousness	7%	3%	1%	0%
Anxiety	5%	1%	4%	1%
Insomnia	5%	1%	5%	1%
Dizziness	5%	4%	5%	2%

Sources: [1][2]. Stopping for an adverse reaction was 8% vs 3% on modafinil (74 of 934) and 7% vs 4% on armodafinil (44 of 645) [1][2]. The labels call headache and anxiety dose-related for modafinil, and headache, rash, depression, dry mouth, insomnia and nausea for armodafinil [1][2].

The UK text classes headache as “very common” (about 21%) and nervousness, insomnia, anxiety, nausea, tachycardia and abnormal liver tests as “common” [3]. A 2026 meta-analysis of 54 studies found risks varying by condition, such as insomnia risk ratios of 5.82 in sleep apnea and 4.97 in ADHD [10].

Serious warnings

Rash and hypersensitivity. The modafinil label (section 5.1) reports that in pediatric trials rash led to discontinuation in about 0.8% (13 of 1,585), including one possible Stevens-Johnson case, versus none among 380 children on placebo. Rare serious cases, including toxic epidermal necrolysis (TEN) and DRESS, were reported in adults and children worldwide. The label says reports of TEN and SJS, though undercounted, exceed the background rate of 1 to 2 cases per million person-years, and that no factor predicts who develops them [1]. Nearly all cases began within 1 to 5 weeks, with isolated cases after about 3 months [1]. The armodafinil label adds skin and mouth sores, blistering, recurrence after rechallenge, and onset from 1 day to 2 months [2]. Both labels describe angioedema; the modafinil label also reports multi-organ hypersensitivity with at least one fatality (median detection 13 days) [1][2]. The UK text lists SJS, DRESS and angioedema as “not known” frequency [3]. The EMA described a “strong” link to serious skin reactions, with higher apparent risk in children [4].

Psychiatric symptoms (section 5.5). Postmarketing reports include mania, delusions, hallucinations, suicidal ideation and aggression, some requiring hospitalisation; many patients had psychiatric histories, not all [1][2]. One healthy volunteer developed paranoid delusions and hallucinations at three times the recommended daily amount while sleep-deprived, with no psychosis 36 hours after stopping [1]. A 2026 review of 24 case reports found hallucinations in 66.7% and delusions in 58.3%, sometimes at therapeutic doses; it cannot give an incidence [11].

Cardiovascular events (section 5.7). The US labels do not recommend use with left ventricular hypertrophy, or mitral valve prolapse syndrome on prior stimulants [1][2]. New or increased antihypertensive use was needed by 2.4% vs 0.7% on modafinil and 2.9% vs 1.8% on armodafinil; armodafinil raised pulse by 0.9 to 3.5 beats per minute on average [1][2]. The UK text contraindicates uncontrolled moderate to severe hypertension and arrhythmia, and recommends an ECG beforehand [3]. The EMA reached the same position in 2011 [4].

Persistent sleepiness (section 5.4). Wakefulness may not return to normal, and patients may not report drowsiness unless asked [1][2].

Interactions as the labels write them

Label statement	Source
Potent CYP3A4/5 inducers (carbamazepine, phenobarbital, rifampin) or inhibitors (ketoconazole, erythromycin) “could alter” modafinil levels	[1], section 12.3
Modafinil is a weak CYP3A inducer; exposure may fall for steroidal contraceptives, cyclosporine (one transplant case: levels down 50%), midazolam, triazolam. Ethinyl estradiol: Cmax down 11%, AUC down 18%; triazolam AUC down 59%	[1], sections 7, 12.3
Contraceptive effectiveness “may be reduced” during use and for one month after (US); the UK text says two months, as did the MHRA in November 2020	[1][2][3][5]
CYP2C19 substrates (phenytoin, diazepam, propranolol, omeprazole, clomipramine) may reach higher exposure; armodafinil raised omeprazole exposure 40%	[1][2]
Warfarin: more frequent INR monitoring “should be considered”; a single-dose study found no significant change, but an interaction “cannot be ruled out”	[1]
MAO inhibitors: “caution should be used”	[1][2]

The UK text adds phenytoin toxicity, regular prothrombin monitoring in the first two months, and lists HIV protease inhibitors, buspirone, calcium channel blockers and statins as CYP3A4/5 substrates [3]. Alcohol with modafinil “has not been studied” [1].

Contraindications and populations with caution

The US contraindication is known hypersensitivity to modafinil or armodafinil (section 4) [1][2]. The UK adds uncontrolled moderate to severe hypertension and arrhythmias, and does not recommend use under 18 [3].

The US labels advise caution with histories of psychosis, depression or mania, a reduced dose in severe hepatic impairment, lower doses and close monitoring in older adults, and note a nine-fold rise in an inactive metabolite in severe renal failure [1][2]. In children, safety and effectiveness are not established; a 6-week trial in 165 children with narcolepsy found no significant benefit, and ADHD trials recorded 3 serious rashes among 933 exposed [1].

Pregnancy and breastfeeding

The US modafinil label still uses “Pregnancy Category C”: no adequate studies in pregnant women, reports of intrauterine growth restriction and spontaneous abortion of unknown cause, and developmental toxicity in rats and rabbits [1]. The armodafinil label says limited data are “insufficient to inform a drug associated risk” and gives the 2 to 4% background malformation rate [2]. The UK text says modafinil “should not be used during pregnancy” or breastfeeding [3]. On milk, the modafinil label says it is unknown whether it passes into human milk; the armodafinil label reports no human data [1][2].

Study	Design	Major malformations
Damkier 2020, Denmark [6]	National registers; 49 exposed vs 828,644 unexposed	6 (12%) vs 3.9%; adjusted OR 2.7 (95% CI 1.1-6.9)
Cesta 2020, Sweden and Norway [7]	133 exposed among 1,917,605 pregnancies	3 (2.3%) vs 2.1%; crude RR 1.06 (0.35-3.25)
Onken 2024, 18 teratology services [9]	175 prospective cases	3 of 150 live births (2.0%); tendency to lower birth weight and head circumference
Kaplan 2025, US registry [8]	191 pregnancies, 2010-2024	18 of 137 prospective live births (13.1%) vs 3% reference

The Danish authors cite residual confounding and small numbers [6]; the Nordic authors note that prescription fills only proxy for use [7]. The registry covered modafinil and armodafinil together; Syneos ran it under contract with Teva, and several authors are Teva employees or consultants [8]. The MHRA’s 2020 notice cited about 15% from a US study of 61 babies [5]. The ENTIS authors called their result preliminary and advised against use in pregnancy [9].

Abuse, dependence and overdose

The modafinil label states it “produces psychoactive and euphoric effects” in humans, similar to methylphenidate in an inpatient study of experienced drug users, and was self-administered by monkeys trained on cocaine [1]. After 9 weeks of use, one trial recorded no withdrawal symptoms over 14 days, though sleepiness returned in narcolepsy [1]. The armodafinil label differs: it reports abuse, diversion and misuse, says physical dependence “can occur”, lists withdrawal symptoms (including shaking, sweating, confusion, aggression and atrial fibrillation) and notes multiple cases of tolerance [2]. The UK text says studies “demonstrated a potential for dependence” [3]. A 2004 review judged abuse potential limited at scale, before later reports [13]. A FAERS analysis of 2019-2023 found 72 modafinil signals from 632 cases, including abuse, restricted fetal growth and spontaneous abortion [12]. FAERS is voluntary reporting, so these are signals, not rates.

Overdose (section 10): trial exposures of 5 to 8 times the recommended daily amount in 32 subjects produced agitation, insomnia and modest haemodynamic rises. Fatal overdoses, alone or with other drugs, have been reported after marketing, and accidental ingestion by children as young as 11 months. There is “no specific antidote”; management is supportive [1][2].

Risks and unknowns

Label rates come from patients with sleep disorders, not healthy people. The healthy-adult literature is mostly single-dose laboratory work; one systematic review saw no excess side effects or mood change [14][15]. We did not find a long-term study in healthy users, or data on how repeated use outside medical care affects skin, psychiatric or cardiovascular risk.

Other gaps: in 2011 the EMA asked for cardiovascular and skin safety studies, a drug-use study and an analysis of student misuse [4]; we did not find the results. Pregnancy findings conflict, and we found no randomized evidence. The label states the wakefulness mechanism is unknown [1].

Frequently asked questions

What are the most common side effects of modafinil?

Headache (34% vs 23% on placebo), nausea (11% vs 3%), nervousness (7% vs 3%), and anxiety and insomnia (each 5% vs 1%) in the US label’s pooled trials of patients with sleep disorders [1].

Does modafinil interact with birth control?

The US label says steroidal contraceptive effectiveness “may be reduced” during use and for one month after stopping, because modafinil induces CYP3A4/5; the UK text and the MHRA say two months [1][3][5].

Can modafinil cause Stevens-Johnson syndrome?

Both US labels list SJS, TEN and DRESS as rare postmarketing events. The modafinil label says reports exceed the background rate and no factor predicts them [1][2].

Is modafinil safe in pregnancy?

The UK text says not to use it in pregnancy, the US label notes no adequate studies, and four studies report malformation rates of 2.0% to 13.1% [1][3][6][7][8][9].

Sources

1. Apotex Corp., “PROVIGIL (modafinil) tablets, prescribing information”, DailyMed (label revised Feb 2025; read as SPL XML 2026-11-08). https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=0391f182-1958-9fef-a944-53b229ce99e7
2. Apotex Corp., “NUVIGIL (armodafinil) tablets, prescribing information”, DailyMed (label text revised Feb 2025; listing published 15 Sep 2026; read as SPL XML 2026-11-08). https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=a69631d0-1263-769e-b232-4ab9160e0fa3
3. Milpharm Ltd, “Modafinil 200 mg tablets, SmPC”, electronic medicines compendium (revised 14 Feb 2024). https://www.medicines.org.uk/emc/product/4320/smpc
4. EMA, “Modafinil - referral” (EMEA/H/A-31/1186; CHMP opinion 18 Nov 2010, Commission decision 27 Jan 2011). https://www.ema.europa.eu/en/medicines/human/referrals/modafinil
5. MHRA, “Sleep disorder drug modafinil linked to increased risk of birth defects and also to reduced effectiveness of contraception” (16 Nov 2020). https://www.gov.uk/government/news/sleep-disorder-drug-modafinil-linked-to-increased-risk-of-birth-defects-and-also-to-reduced-effectiveness-of-contraception
6. Damkier P, Broe A, “First-trimester pregnancy exposure to modafinil and risk of congenital malformations”, JAMA 2020;323:374-6. https://jamanetwork.com/journals/jama/fullarticle/2759460
7. Cesta CE et al., “Incidence of malformations after early pregnancy exposure to modafinil in Sweden and Norway”, JAMA 2020;324:895-7 (erratum JAMA 2021;325:90, not read). https://jamanetwork.com/journals/jama/fullarticle/2769999
8. Kaplan S et al., “Pregnancy and fetal outcomes following prenatal exposure to modafinil and/or armodafinil: a 14-year registry study”, Neurol Clin Pract 2025;15:e200551. https://pubmed.ncbi.nlm.nih.gov/41070135/
9. Onken M et al., “Effects of maternal modafinil treatment on fetal development and neonatal growth parameters (ENTIS)”, Acta Psychiatr Scand 2024;150:372-84. https://pubmed.ncbi.nlm.nih.gov/38110225/
10. Jung J et al., “Assessing condition-specific adverse event profiles of modafinil for labelled and off-label uses”, Basic Clin Pharmacol Toxicol 2026;138:e70147. https://pubmed.ncbi.nlm.nih.gov/41367108/
11. Sato Y et al., “Modafinil-associated psychosis: a systematic review of case characteristics and management strategies”, Ther Adv Psychopharmacol 2026;16. https://pubmed.ncbi.nlm.nih.gov/42137740/
12. Zhou X et al., “Evaluation of pitolisant, sodium oxybate, solriamfetol, and modafinil for the management of narcolepsy: a retrospective analysis of the FAERS database”, Front Pharmacol 2024;15:1415918. https://pubmed.ncbi.nlm.nih.gov/39588154/
13. Myrick H et al., “Modafinil: preclinical, clinical, and post-marketing surveillance - a review of abuse liability issues”, Ann Clin Psychiatry 2004;16:101-9. https://pubmed.ncbi.nlm.nih.gov/15328903/
14. Battleday RM, Brem AK, “Modafinil for cognitive neuroenhancement in healthy non-sleep-deprived subjects: a systematic review”, Eur Neuropsychopharmacol 2015;25:1865-81. https://pubmed.ncbi.nlm.nih.gov/26381811/
15. Roberts CA et al., “How effective are pharmaceuticals for cognitive enhancement in healthy adults?”, Eur Neuropsychopharmacol 2020;38:40-62. https://pubmed.ncbi.nlm.nih.gov/32709551/

This article is information, not medical advice.