The purpose of this study was to determine the optimal bolus dose of remifentanil required for the successful insertion of the laryngeal mask airway during propofol induction in children without a neuromuscular blocking agent. Twenty-six paediatric patients, aged 3-10 years, requiring anaesthesia for short ambulatory surgery were recruited. A predetermined bolus dose of remifentanil was injected over 30 s, followed by propofol 2.5 mg.kg(-1) over 10 s. The bolus dose of remifentanil was determined by a modified Dixon's up-and-down method, starting from 0.5 microg.kg(-1) (0.1 microg.kg(-1) as a step size). Laryngeal mask insertion was attempted 90 s after the end of remifentanil injection and the response of patients was classified as either 'movement' or 'no movement'. The bolus dose of remifentanil at which there was a 50% probability of successful laryngeal mask insertion (ED(50)) during induction with 2.5 mg.kg(-1) propofol was 0.56 (0.07) microg.kg(-1) in children without a neuromuscular blocking agent. From probit analysis, the ED(50) and ED(95) of remifentanil were 0.52 microg.kg(-1) (95% confidence limits, 0.42-0.62 microg.kg(-1)) and 0.71 microg.kg(-1) (95% confidence limits, 0.61-1.40 microg.kg(-1)), respectively.