Using the National Ambulatory Medical Care Survey, a nationally representative sample of ambulatory visits, we examined modafinil use from January 1, 2002, through December 31, 2009.3 We categorized patients by whether or not they had an on-label indication for modafinil, defining off-label use as the absence of an on-label diagnosis, and examined the association between modafinil use and specific off-label indications.1 We further examined modafinil use among patients treated by psychiatrists, neurologists, pulmonologists, and otolaryngologists relative to primary care physicians and other specialists. Finally, we examined the association between modafinil use and other centrally acting medications that might interact with modafinil: antidepressants, benzodiazepines, and amphetamines. In multivariate logistic regression, modafinil use was the outcome, and the predictor variables of interest were, in separate models, diagnostic indication, physician specialty, and concurrent medication. Analyses controlled for age, sex, race/ethnicity, payer, geographic region, and survey year. Analyses were survey weighted and performed using statistical software (Stata/IC, version 12; StataCorp). A supplemental description of our methods is available online (eMethods).