The clinical pharmacology of fentanyl and alfentanil was examined in opioid-experienced volunteers with agonist and antagonist sensitivity measures. Two studies used within-subject, placebo-controlled, crossover designs. In study 1, fentanyl (0.125, 0.25 mg/70 kg i.v.) was followed at 0, 20, 60 and 180 min by naloxone (10 mg/70 kg i.m.). Agonist effects during 180-min and 0-min (control; simultaneous fentanyl-naloxone i.v. infusion) challenge sessions were compared. Fentanyl rapidly constricted pupils, depressed respiration and produced subjective "high" and opiate symptoms lasting 60 to 120 min, depending on the measure. Naloxone precipitated withdrawal symptoms of comparable intensity at each challenge point. In study 2, fentanyl (0.125, 0.25 mg/70 kg i.v.), alfentanil (1, 2 mg/70 kg i.v.) and saline were followed at 1 and 6 hr by naloxone (10 mg/70 kg i.m.). Agonist effects were examined during 6-hr challenge sessions. The two drugs produced a comparable range of effects with similar peak magnitude for 0.125 mg/70 kg fentanyl and 1 mg/70 kg alfentanil and for 0.25 mg/70 kg fentanyl and 2 mg/70 kg alfentanil. Alfentanil's duration of action was brief ( < 60 min). Withdrawal was precipitated at 6 hr only after 0.25 mg/70 kg fentanyl. These findings support typical mu opioid characteristics (pleasurable subjective effects, physical dependence) for both drugs, differential duration of action (fentanyl > alfentanil) and peak effects consistent with a 1:8 (fentanyl/alfentanil) potency ratio.