Two groups of children, aged 1-4 years (n = 28) and 5-10 years (n = 28), respectively, received at random one of four doses of rocuronium (0.12, 0.17, 0.22 or 0.27 mg kg(-1)). When maximum block was obtained, further rocuronium to a total dose of 0.5 mg kg(-1) was given. At a spontaneous T1 recovery of 25% the block was reversed with atropine and neostigmine in half the patients. The remainder were allowed to recover spontaneously. There was no difference in potency in the two age groups. An ED(50) of 0.2 mg kg(-1) was estimated. The estimates of ED(90) were model-dependent of approximately 0.32 mg kg(-1). The maximum block was found significantly higher in the younger age group (99.0 +/- 1.5% (mean +/- SD)) as compared to the older group (97.5 +/- 2.3%), and the clinical duration was also longer (16.6 +/- 5.3 min vs. 13.3 +/- 3.8 min), respectively. There was no significant difference between the two age groups in duration(90) (26.9 +/- 6.5 min, 22.5 +/- 6.7 min, respectively) and duration(0.7) (27.6 +/- 6.0 min, 24.9 +/- 7.9 min, respectively). Recovery time(25-75), recovery time(25-90), but not recovery time(25-0.7) were found significantly longer in the 1-4 year group as compared to the times in older children. Neostigmine administration reduced recovery time by approximately half to two-thirds. MAP was not influenced by rocuronium. Following the injection of rocuronium in the younger age group there was a 15% increase in heart rate compared to a 10% increase in the age group 5-10 years.