Forty-four male, neuroleptic-free, acutely psychotic patients with at least one diagnosis of schizophrenia among 11 diagnostic systems, and 28 healthy controls, underwent measurement of prolactin (PRL) concentrations before and after intravenous administration of haloperidol (0.5 mg). Basal PRL concentrations were lower in the patients with Research Diagnostic Criteria (RDC) DSM-III, Cloninger, and Taylor and Abrams schizophrenias than in controls. Compared with the controls, the PRL response to haloperidol was lower in the patients with schizophrenia defined by all diagnostic systems except those of Schneider and M. Bleuler. Neither basal nor stimulated PRL concentrations were correlated with positive symptoms, but basal PRL was correlated with the Brief Psychiatric Rating Scale (BPRS) depression-related subscore. This study lends further support for the presence of dopaminergic dysfunction in schizophrenia, and demonstrates the advantages and problems in the use of multidiagnostic psychopathological evaluation to categorize a disorder where there is major disagreement among diagnostic systems.