The aim of this study was to compare recovery and psychomotor performance after maintenance of anaesthesia with sevoflurane or sevoflurane supplemented with remifentanil. Sixty-six per cent nitrous oxide was used in all patients. Twenty patients each were randomly allocated to maintenance of anaesthesia with sevoflurane only in concentrations necessary to maintain adequate anaesthesia or with 1.5, 1.0 or 0.5 MAC (end-tidal) of sevoflurane supplemented with remifentanil. The median dosage of remifentanil required in the last three groups was 0.21, 0.25 and 0.34 microg x kg(-1) x min(-1), respectively (p < 0.05). The median times to eye opening were 10.3, 12.7, 11.0 and 6.5 min in the four groups (p < 0.05 between the 0.5 MAC and the other groups) and for orientation 12.1, 14.9, 12.3 and 8.3 min, respectively (p < 0.05 between 0.5 and 1.5 MAC groups). There was no significant difference in the mini-mental state assessment scores or the actual discharge times from the recovery ward among the groups. Significantly greater numbers of patients could perform the critical flicker fusion test at 15 min in the group receiving the lowest concentration of sevoflurane and the highest dosage of remifentanil (p < 0.05). Patients in this group also showed the highest incidence of chest wall rigidity (p < 0.003). We conclude that, while the use of remifentanil with lower concentrations of sevoflurane facilitates early recovery, it does not influence discharge time from recovery ward and may be associated with side-effects such as chest wall rigidity.