The combination of vecuronium and atracurium was studied in 60 children of ASA physical status 1 or 2. In part I, the dose-response relationships were determined in 30 children who were randomly assigned to receive a single bolus of 0.02, 0.025 or 0.03 mg kg-1 of vecuronium or 0.075, 0.01 or 0.0125 mg.kg-1 of atracurium. The evoked electromyogram of the adductor pollicis brevis muscle to train-of-four stimulation was monitored. The speed of onset of the neuromuscular blockade was determined by a 95% depression of the train-of-four whereas the recovery index was established at 75% recovery of T1. In part II, 30 children were randomly assigned to receive 2 x ED95 of vecuronium, 2 x ED95 of atracurium, or a combination of 1 x ED95 of vecuronium + 1 x ED95 of atracurium. This study showed that the ED50 and ED95 for vecuronium were 0.021 mg.kg-1 and 0.037 mg.kg-1 and for atracurium 0.11 mg.kg-1 and 0.30 mg.kg-1, respectively. The slopes of the dose-response relationships were significantly different (p < 0.001). With vecuronium alone, the speed of onset of neuromuscular blockade was significantly slower (p < 0.001) and the duration of action less (p < 0.001) than that with atracurium alone or with the combination. There were no differences between atracurium and the combination of both medications. The recovery index was similar for all groups. We conclude that the dose-response relationships of vecuronium and atracurium in children undergoing halothane anaesthesia are not parallel and the neuromuscular effects of vecuronium and atracurium are neither additive nor synergistic. While vecuronium has a shorter duration of action than atracurium, this feature is not apparent when it is combined with atracurium in equipotent doses. Recovery is rapid and not prolonged when these two drugs are combined.