The antipsychotic effect of neuroleptics cannot be attributed entirely to acute blockade of postsynaptic D(2)-like dopamine (DA) receptors, but may arise in conjunction with the delayed depolarization block of the presynaptic neurons and reduced DA synthesis capacity. Whereas the phenomenon of depolarization block is well established in animals, it is unknown if a similar phenomenon occurs in humans treated with neuroleptics. We hypothesized that haloperidol treatment should result in decreased DA synthesis capacity. We used 6-[(18)F]fluoro-L-dopa (FDOPA) and positron emission tomography (PET) in conjunction with compartmental modeling to measure the relative activity of DOPA decarboxylase (DDC) (k(D)(3), min(-1)) in the brain of nine unmedicated patients with schizophrenia, first in the untreated condition and again after treatment with haloperidol. Patients were administered psychometric rating scales at baseline and after treatment. Consistent with our hypothesis, there was a 25% decrease in the magnitude of k(D)(3) in both caudate and putamen following 5 weeks of haloperidol therapy. In addition, the magnitudes of k(D)(3) in cerebral cortex and thalamus were also decreased. Psychopathology as measured with standard rating scales improved significantly in all patients. The decrease of k(D)(3) in the thalamus was highly significantly correlated with the improvement of negative symptoms. Subchronic treatment with haloperidol decreased the activity of DDC in the brain of patients with schizophrenia. This observation is consistent with the hypothesis that the antipsychotic effect of chronic neuroleptic treatment is associated with a decrease in DA synthesis, reflecting a depolarization block of presynaptic DA neurons. We link an alteration in cerebral catecholamine metabolism in human brain with the therapeutic action of neuroleptic medication.