The onset, maximal neuromuscular block, and duration of rocuronium were compared with atracurium and vecuronium during enflurane anesthesia. Sixty patients received rocuronium (80, 100, 120, or 160 micrograms/kg). Enflurane enhanced a rocuronium neuromuscular block in a dose-related manner; the ED50 was 104 +/- 11 and 83 +/- 7 micrograms/kg (SEM) during 1% and 2% enflurane anesthesia, respectively. Patients receiving atracurium (0.12 mg/kg) or vecuronium (0.02 mg/kg) were studied during 1% enflurane anesthesia until seven in each group qualified by achieving a maximal block between 85% and 97%. These patients were matched with each other and with patients who had received rocuronium. Seven groups of three patients (rocuronium, vecuronium, and atracurium) were obtained. The average difference in maximal block was less than 2% between matched patients. The ratio of dose used to achieve a similar final block suggests potency ratios of 1, 8.5, and 1.2 for rocuronium, vecuronium, and atracurium. Rocuronium's onset time (time from drug administration to 50%, 75%, and 90% of final block) was significantly faster than either of the other two muscle relaxants (P < 0.01). Time to 90% of final block was 1.35 min for rocuronium, 3.06 min for atracurium, and 3.71 min for vecuronium. Using these equipotent doses, atracurium also had a shorter time to develop neuromuscular block than vecuronium (P < 0.05). For these three intermediate duration neuromuscular blockers, speed of onset was inversely related to their potency, confirming a relationship that had been demonstrated for the long-acting drugs pancuronium, d-tubocuranine, and gallamine.