We assessed the safety and efficacy of modafinil for the treatment of excessive daytime sleepiness in patients with Parkinson's disease (PD). This was a single-site, randomized, double-blind, placebo-controlled crossover study of 21 PD patients having an Epworth Sleepiness Scale (ESS) score > or =10. They received either placebo or modafinil 200 mg/day for 3 weeks, followed by a washout week, then the alternate treatment for 3 weeks. The ESS data demonstrated a carryover effect, so the changes from baseline ESS scores were compared between the two treatments for period 1 only. The ESS scores for the placebo group went from 16.0 +/- 4.2 (mean +/- SD) to 17.0 +/- 5.1 and for the modafinil group went from 17.8 +/- 4.2 to 14.4 +/- 5.7 (P = 0.039). There was no significant carryover effect for any other measure. The patient Clinical Global Impression of Change (+3 to -3) improved by 0.75 on modafinil compared with 0.15 for placebo (P = 0.07). A total of 7 of 20 (35%) of the patients reported some improvement on modafinil but not placebo. There was no significant improvement or worsening of the UPDRS subscores I-III, Timed Tap test, or time on. Vital signs, electrocardiograms, and lab tests were unchanged. Modafinil was very well tolerated. Our data demonstrate that, in a small sample size, administration of 200 mg/day of modafinil was associated with few side effects and was modestly effective for the treatment of excessive daytime sleepiness in patients with PD.