Twenty patients were randomly anaesthetized with either thiopental 5 mg/kg followed by a 15 mg/kg/h continuous infusion, or propofol 2.5 mg/kg followed by a 9 mg/kg/h continuous infusion, paralysed with vecuronium 0.1 mg/kg, intubated and ventilated with nitrous oxide 50% in oxygen. Fifteen minutes after induction, fentanyl 5 micrograms/kg was injected. Inspiratory tracheal pressure (PT), gas flow (V) and volume (V) were continuously measured while the lungs were inflated with a constant inspiratory flow ventilator. Respiratory compliance (Crs) and resistance (Rrs) were calculated from the regression of PT on V. In both groups Crs decreased following anaesthesia. Fentanyl injection elicited an increase in Rrs (from 1.04 +/- 0.70 to 1.63 +/- 0.92 kPa x l-1 x s) and a further decrease in Crs (from 0.55 +/- 0.30 to 0.42 +/- 0.10 l x kPa-1) in the thiopental group but not in the propofol group (Rrs: 1.26 +/- 0.69 to 1.08 +/- 0.44 kPa x l-1 x s, Crs: 0.49 +/- 0.11 to 0.48 +/- 0.13 l x kPa-1). These results suggest that the dose of propofol administered in this study may prevent fentanyl-induced bronchoconstriction.