The authors have previously attributed the mechanism for the 50-67% decrease in the required dose of thiopental for induction of anesthesia in aged human patients to a decrease in the initial distribution volume for thiopental. Using a larger group of patients and volunteers studied in the laboratory, the authors have re-examined thiopental pharmacokinetics and EEG pharmacodynamics relative to age. A population data analysis approach (NONMEM), using a three-compartment model, was used to analyze bolus and rapid iv infusion thiopental serum concentration versus time data from 64 subjects. A one-compartment model was also used on the first 10 min of serum concentration data to focus only on the initial distribution phase. The population pharmacokinetic analysis demonstrated that when thiopental is administered via an iv bolus injection, traditional pharmacokinetic models limit the accurate characterization of thiopental distribution phenomena. Using the rapid iv infusion data, the pharmacokinetic mechanism for the decreased thiopental dose requirement in the elderly was a decreased rapid intercompartment clearance. Thiopental distribution from the central compartment of the three-compartment model to the rapidly equilibrating compartment (rapid intercompartment clearance) decreased 27% between the ages of 35-80 yr and decreased 34% in the one-compartment analysis. EEG spectral edge versus time data from 37 subjects was analyzed with a semiparametric modelling approach to remove the disequilibrium between thiopental serum concentration and the spectral edge. A population data analysis (NONMEM) was performed with several pharmacodynamic models. There was no age-related change in brain responsiveness or pharmacodynamics when the spectral edge is used as a measure of drug effect.