In this paper we have described early applications of computerized EEG techniques in psychopharmacology. Perhaps our most remarkable finding was there were practically no differences between very chronic drug free schizophrenic patients and normals, which contradicts much of the EEG imaging literature. To us, the most likely explanation is that most of the anterior slowing observed in other studies was due to contamination from orbital artifacts, which we took exceptional pains to remove. Lingering effects of neuroleptic medications may also have contributed. Alternatively, EEG deviations in schizophrenia may recede when the illness reaches a very chronic stage, although this hypothesis is less tenable. There were significant differences between placebo and the three neuroleptics in terms of increased amplitudes in the delta and theta frequency bands in the anterior head regions, which is compatible with data from other studies. These changes were most pronounced with clozapine and least prominent with haloperidol, with chlorpromazine occupying an intermediate position. This order happens to parallel their relative antiserotonergic, antihistaminic and anticholinergic properties. The latter may have been partially obscured by the addition of benztropine. In a subgroup of patients who were recorded under each of the treatment conditions, there were more fast frequencies with clozapine than with the other neuroleptics agreeing with Roubicek and Major. This could be a function of clozapine's increased adrenergic activity as reported by Ackenheil. An unexpected finding was that patients who responded to clozapine had higher amplitudes in the alpha spectrum, most pronounced in the left anterior quadrant, than did the nonresponders. These differences between responders and nonresponders obtained whether patients were on placebo, haloperidol or clozapine. Curiously, Buchsbaum et al. found that anxious patients who responded to benzodiazepines also had higher alpha amplitudes in the same brain regions, which differentiated them from nonresponders. These findings clearly warrant future scientific investigation. In this regard, the generalizability of our data is limited by the extremely chronic, treatment-resistant population studied. However, promising directions for further research in EEG and psychopharmacology have been identified.