The effects of a number of psycho-active drugs on electrocorticographic activity (ECoG) of the rat have been studied. Frontoparietal ECoG was recorded during 6-min periods immediately before drug/vehicle administration and starting at 20 and 45 min after administration. For each 6-min recording time, a mean power spectrum was calculated and smoothed. A quotient of the power at 20 and 45 min after and before drug, respectively, was then calculated. These quotients were used as input for an analysis of variance, followed by a t test and a normalization for the degrees of freedom, resulting in so-called n-profiles. Although each drug has its own characteristic n-profile, n-profiles of drugs belonging to the same clinical class have some characteristics in common. An automated classification system of psychotropic drugs, based on parameters extracted from n-profiles by discriminant analysis, is presented. For the antidepressant drug class the success rate of correct classification of antidepressant drugs is between 53 and 88%, for a neuroleptic drug in the neuroleptic drug class the success rate is 87%, for an anxiolytic drug in the anxiolytic drug class 100% and for a psychostimulant drug in the psycho-stimulant drug class between 86 and 100%. For a reliable classification of a drug about 10 n-profiles of active doses are needed. Using n-profiles, it appeared also possible to discriminate between antidepressant, neuroleptic, anxiolytic and psychostimulant drugs based on visual judgement. Moreover, visual evaluation of the n-profiles enables 4 subclasses to be recognized within the antidepressant class, 2 subclasses within the neuroleptic and 2 subclasses within the anxiolytic class.