Scopolamine is a muscarinic receptor antagonist commonly used as a pharmacological model substance based on the "cholinergic hypothesis" of memory loss in senile dementia of the Alzheimer type. The objective of the study was to relate pharmacodynamic electroencephalogram (EEG) changes and scopolamine serum concentration using pharmacokinetic-pharmacodynamic (PK-PD) modeling techniques. This was a randomized, three-way crossover, open-label study involving 10 healthy nonsmoking young male volunteers who received either scopolamine 0.5 mg as an intravenous (i.v.) infusion over 15 minutes or an intramuscular (i.m.) injection or a placebo. The pharmacodynamic EEG measure consists of the total power in delta, theta, alpha, and beta bands over frontal, central, and occipital brain areas. The values of the pharmacokinetic parameters of scopolamine after i.v. infusion were clearance (CL) 205 +/- 36.6 L/h, volume of distribution (Vd) 363 +/- 66.7 L, distribution half-life (t1/2 alpha) 2.9 +/- 0.67 min, and terminal half-life (t1/2 beta) 105.4 +/- 9.94 min (mean +/- SEM). Mean peak serum concentrations (Cmax) were 4.66 and 0.96 ng/ml after i.v. and i.m. administration, respectively (p < 0.05). The area under the serum concentration versus time curve (AUC) after i.m. administration (81.27 +/- 11.21 ng/ml/min) was significantly lower compared to the value after i.v. infusion (157.28 +/- 30.86 ng/ml/min) (mean +/- SEM, p < 0.05). Absolute bioavailability of scopolamine after i.m. injection was 57% +/- 0.08% (mean +/- SEM). After both i.v. and i.m. administration, scopolamine induced a decrease in EEG alpha power (7.50-11.25 Hz) over frontal, central, and occipital brain areas compared to placebo (p < 0.05). The individual concentration-EEG effect relationships determined after i.v. infusion of scopolamine were successfully characterized by a sigmoidal Emax model. The averaged values of the pharmacodynamic parameters were E0 = 0.58 microV2, Emax = 0.29 microV2, EC50 = 0.60 ng/ml, and gamma = 1.17. No time delay between serum concentrations and changes in alpha power was observed, indicating a rapid equilibration between serum and effect site. The results provide the first demonstration of a direct correlation between serum concentrations of scopolamine and changes in total power in alpha frequency band in healthy volunteers using PK-PD modeling techniques. As regards the effect on the EEG, 0.5 mg of scopolamine administered i.v. appears to be a suitable dose.