This study analyzed interrelationships among plasma level of haloperidol (HAL), electroencephalographic (EEG) changes, and clinical response in 37 acutely exacerbated schizophrenic patients after a 6-week period of treatment. Two hypotheses were tested: (1) EEG theta response to HAL depends on levels of HAL in plasma, and this relationship is expressed in patients showing a clear clinical response (responders). (2) Responders and nonresponders are characterized by a different neuroleptic EEG profile. EEG examinations (resting, waking EEG) were performed at study entry, end point of the placebo period ("baseline"), and weekly during the entire HAL treatment period. EEG response was measured by power spectral changes in four frequency bands (delta, theta, alpha, and beta); clinical response was assessed by the Brief Psychiatric Rating Scale. There was a significant relationship between HAL plasma levels and EEG theta activity for treatment responders, whereas no relationship was detected for the nonresponders. Furthermore, there were EEG changes (in the delta and alpha bands) that depended on clinical response but did not show any relationship, either in responders or nonresponders, to HAL plasma levels. These results supported both hypotheses.