OBJECTIVE Pharmacokinetics of propofol in rats have usually been described using linear models. Furthermore, there are only a few investigations for a pharmacodynamic model of the electroencephalographic effects of propofol in rats. We investigated pharmacokinetics and pharmacodynamics of propofol in rats with special regard to linearity in pharmacokinetics and development of tolerance. METHODS Twelve adult male Sprague-Dawley rats received propofol in three successive infusion periods of 30 min each with infusion rates of 0.5, 1 and 0.5 mg kg(-1) min(-1). Propofol plasma concentrations were determined from arterial blood samples. Pharmacokinetics were tested for linearity using the ratio of the concentrations at the end of the first and second infusion interval as a model independent criterion. Several linear and nonlinear models were investigated with population pharmacokinetic analysis. Pharmacodynamics were analysed using the median frequency of the electroencephalographic power spectrum as a quantitative measure of the hypnotic effect. RESULTS Pharmacokinetics were found to be nonlinear and were best described by a two-compartment model with Michaelis-Menten elimination (Vm = 2.17 microg mL(-1) min(-1), Km = 2.65 microg mL(-1), k12 = 0.30 min(-1), k21 0.063 min(-1), Vc = 0.13 L). Acute tolerance to the electroencephalographic effect of propofol was observed. The hypnotic effect was best described by a sigmoid Emax model (E0 = 17.8 Hz, Emax = 17.7 Hz, EC50 = 4.1 microg mL(-1), gamma = 2.3, ke0 = 0.36 min(-1)) with competitive antagonism of propofol and a hypothetical drug in an additional tolerance compartment. CONCLUSIONS For the applied infusion scheme, propofol pharmacokinetics in rats were nonlinear and a development of tolerance to the electroencephalographic effect of propofol was observed during an infusion time of 90 min.