BACKGROUND Mivacurium's rapid onset and short duration of action in children suggests that intramuscular administration might treat laryngospasm and facilitate tracheal intubation without producing prolonged paralysis. Accordingly, the authors measured the neuromuscular effects of intramuscular mivacurium in anesthetized infants and children. METHODS Twenty unpremedicated infants and children (3 months to 5 yr of age) were anesthetized with nitrous oxide and halothane and permitted to breathe spontaneously. When anesthetic conditions were stable, mivacurium was injected into the quadriceps or deltoid muscle. Minute ventilation and adductor pollicis twitch tension were measured. The initial mivacurium dose was 250 micrograms/kg and was increased (to a maximum of 800 micrograms/kg, at which dose the trial was ended) or decreased according to the response of the previous patient, the goal being to bracket the dose producing 80-90% twitch depression within 5 min of drug administration. RESULTS No patient achieved > 80% twitch depression within 5 min of mivacurium administration. Peak twitch depression was 90 +/- 13% (mean +/- SD) for infants and 88 +/- 15% for children at 15.0 +/- 4.6 min and 18.4 +/- 6.4 min, respectively. Ventilatory depression (a 50% decrease in minute ventilation or a 10-mmHg increase in end-tidal carbon dioxide tension) occurred at 9.0 +/- 4.4 min in nine infants and 13.6 +/- 7.5 min in 10 children; ventilatory depression did not develop in one infant given a dose of 350 micrograms/kg. Time to peak twitch depression or ventilatory depression was not faster with larger doses. CONCLUSIONS Although ventilatory depression preceded twitch depression, both occurred later with intramuscular mivacurium than would be expected after intravenous mivacurium or intramuscular succinylcholine. The authors speculate that the onset of intramuscular mivacurium is too slow to treat laryngospasm or to facilitate routine tracheal intubation in infants or children, despite administration of large doses.