In the present 6-week double-blind, randomised, multicentre study, the atypical neuroleptic remoxipride was compared to haloperidol in acute schizophrenic patients (DSM-III). Seventy-one patients entered the study, 36 in the remoxipride group and 35 in the haloperidol group. There were ten early withdrawals, four in the remoxipride group and six patients in the haloperidol group. The Present State Examination (PSE) profile revealed a similar reduction in the symptom clusters of psychosis in both treatment groups. Forty-seven per cent of the patients in the remoxipride group and 34% of the patients in the haloperidol group showed clinically relevant improvement (reduction of BPRS total score greater than or equal to 50%). All extrapyramidal symptoms except "glabella tap" occurred significantly less frequently in the remoxipride group as compared to the haloperidol group. Substantially lower incidences of EPS were found by active questioning in the remoxipride group compared to the haloperidol group. In addition, considerably lower incidences were observed in the remoxipride group with respect to drowsiness/somnolence, tiredness/fatigue and concentrating difficulty. At the end of treatment 66% of the patients in the haloperidol group and 22% in the remoxipride group were using anticholinergics. No consistent changes were found in the mean plasma HVA level in either treatment group. In responders (reduction of BPRS total score greater than or equal to 50%) lower baseline HVA levels were observed in both treatment groups. This study indicates that the newly developed neuroleptic remoxipride is an effective antipsychotic compound, which is clinically safe and well tolerated. In particular, few EPS were induced by remoxipride, as compared to haloperidol.