Using power spectral analysis of the electroencephalogram (EEG) to measure the effect of thiopental on the brain, the authors investigated the phenomenon of acute tolerance. Three sequential infusions of thiopental, 20–25 min apart, were given to eight healthy volunteers. The infusions were stopped when moderately deep anesthesia, indicated by burst-suppression on the EEG, was reached. The mean (±SD) doses of thiopental for the first, second, and third infusions were 9.6 ± 2.0, 5.6 ± 0.9, and 5.2 ± 1.2 mg·kg-1, respectively. The spectral edge (Hz), defined as the frequency below which 95% of the total EEG power is located, was used to measure thiopental effect. A pharmacodynamic model was used to quantify the relationship of the plasma concentration of thiopental to its effect on the spectral edge. The model estimates the baseline spectral edge, Emax (Hz), the maximal decrease of the spectral edge due to thiopental, Emax (Hz), and the thiopental serum concentration required to produce 50% of the maximal shift of the spectral edge, the IC50 (μg·ml-1). The IC50 is an index of brain sensitivity to thiopental. If acute tolerance to thiopental had developed, the IC50 of the second and third infusions would have been greater than the IC50 of the first infusion. However, there were no significant differences between the values of the IC50 of each infusion (15.9 ± 5.1, 13.9 ± 3.4, and 16.0 ± 4.4 μg·ml-1 respectively), indicating that acute tolerance did not develop during repeated infusions of thiopental. The values for Emax and Emax also did not change significantly, providing additional evidence that the concentration-effect relationship remained constant. The combination of power spectral analysis of the EEG with pharmacodynamic modeling may prove to be a powerful tool for studying the clinical pharmacology of intravenous anesthetics.