Data from animal studies indicate neuroleptic drugs act via their properties as antagonists of CNS dopamine (DA) receptors and this finding has led to the suggestion that alterations in CNS DA neuronal function are associated with psychotic disorders. Clinical investigations of this hypothesis, however, have been hindered by the lack of the availability of a direct and relatively easily obtained index of CNS DA neuronal activity. The work reported here was aimed at the development of such an index. Using a double blind design, human male subjects were given either placebo or debrisoquin, which is a monoamine oxidase inhibitor which does not penetrate brain. On the baseline day (no debrisoquin) and after 6 and 13 days of drug administration blood samples were obtained. In addition, for some patients CSF specimens were obtained via lumbar puncture on the baseline day and after 13 days of drug administration. It was found that debrisoquin produced a highly significant decrease in plasma homovanillic acid (HVA) concentrations whereas the concentrations of HVA in CSF were unchanged. In addition, it was found that the correlation between CSF and plasma HVA prior to debrisoquin was non-significant (r = 0.39, p = N.S., N = 10) whereas after 13 days of debrisoquin treatment the correlation was highly significant (r = 0.95, p less than .01, N = 7). These findings suggest that the administration of debrisoquin produces a situation in which plasma HVA reflects CNS HVA production, and as such debrisoquin may be a useful tool for the clinical investigator who is interested in studying relationships in human subjects between CNS DA neuronal system function and psychopathological states or other disorders which may be mediated via brain DA systems.