The objective of this investigation was to characterize quantitatively the influence of the rate of increase in blood concentrations on the pharmacodynamics of midazolam in rats. The pharmacodynamics of midazolam were quantified by an integrated pharmacokinetic-pharmacodynamic modelling approach. Using a computer controlled infusion technique, a linear increase in blood concentrations up to 80 ng ml(-1) was obtained over different time intervals of 16 h, resulting in rates of rise of the blood concentrations of respectively, 1.25, 1.00, 0.87, 0.46, 0.34 and 0.20 ng ml(-1) min(-1). In one group of rats the midazolam concentration was immediately brought to 80 ng ml(-1) and maintained at that level for 4 h. Immediately after the pretreatment an intravenous bolus dose was given to determine the time course of the EEG effect in conjunction with the decline of midazolam concentrations. The increase in beta activity (11.5-30 Hz) of the EEG was used as pharmacodynamic endpoint. For each individual animal the relationship between blood concentration and the EEG effect could be described by the sigmoidal Emax model. After placebo, the values of the pharmacodynamic parameter estimates were Emax = 82+/-5 microV, EC50,u = 6.4+/-0.8 ng ml(-1) and Hill factor = 1.4+/-0.1. A bell-shaped relationship between the rate of change of midazolam concentration and the value of EC50,u was observed with a maximum of 21+/-5.0 ng ml(-1) at a rate of change of 0.46 ng ml(-1) min(-1); lower values of EC50,u were observed at both higher and lower rates. The findings of this study show that the rate of change in plasma concentrations is an important determinant of the pharmacodynamics of midazolam in rats.