In an acute, double-blind placebo-controlled study, three groups of nine healthy subjects were included. Each group received, in randomized and weekly intervals, a single oral dose of dexfenfluramine (15 mg, 30 mg or 60 mg) or placebo. Psychometric and EEG studies were carried out before as well as 1, 2, 3, 4, 5 and 6 h after drug administration. Changes from predrug to postdrug conditions for each time were determined by the ratio of absolute spectral power and compared to the evolution of power spectra under placebo conditions. Statistical evaluation was done with ANOVA. Dexfenfluramine induced a dose-dependent decrease of power spectra for theta and alpha 1 bands and an increase for the beta band. Topographic brain mapping of these significant changes displayed a central and posterior decrease for theta and alpha 1 bands, and a temporal localization for the beta-power. Maps of relative power-enhanced changes were seen in absolute power and provided false displays. Psychometric evaluation of dexfenfluramine effects only showed trends to extraversion and increased mood scores but no statistical significance was found. Arousal and performance tests were unchanged. These results suggest that qEEG variations are more sensitive to serotonergic drug effects than psychometric investigations. It appears that spontaneous EEG power spectra variations with time must be accounted for by drug evaluation. Absolute power variations are more reliable than relative power. Methodological implications are discussed.