Fentanyl is a mu opiate agonist which is occasionally abused by medical personnel who have ready access to the drug. We examined in healthy volunteers (N = 13) the subjective and psychomotor-impairing effects of intravenous fentanyl (0-100 micrograms/70 kg). A randomized, placebo-controlled, crossover design was used in which subjects were injected with 0, 25 (N = 6), 50 and 100 micrograms/70 kg fentanyl in a double-blind fashion. Subjects completed several questionnaires commonly used in abuse liability testing studies before drug injection and at periodic intervals for up to 3 h after drug injection. Subjects also completed several psychomotor tests at these times. Some aspects of psychomotor functioning (e.g., eye-hand coordination) were impaired by fentanyl. Fentanyl produced dose-related increases in ratings of "high" and "sedated," but also tended to produce dysphoria and somatic symptomatology. Most subjects reported liking the effects of the two higher doses of fentanyl for at least a brief time after injection, but they varied widely in their linking ratings across the 3-h post-drug injection period. Despite the transient increases in liking ratings, fentanyl did not increase scores on a widely-used measure of drug-induced euphoria (morphine-benzedrine group scale of the Addiction Research Center Inventory). The present results suggest that some medical personnel who experiment with fentanyl may like it, and thus be at increased risk for abusing the drug in the future.