We carried out a four-week double-blind placebo-controlled study comparing remoxipride (n = 20) to chlorpromazine (n = 21) and placebo (n = 21) in the treatment of newly admitted schizophrenic patients with acute exacerbation. Chlorpromazine was found to be significantly better than remoxipride on the dropout rate due to inefficacy, Clinical Global Impression (CGI) of severity of illness and Brief Psychiatric Rating Scale (BPRS). Chlorpromazine tended to be better than placebo on the dropout rate related to inefficacy, Nurse's Global Impression (NGI) of severity and on the BPRS measures of positive symptoms (hallucinatory behaviour and thinking disturbance factor). We were unable to detect a difference between remoxipride and placebo except that remoxipride was better in patients who had previously responded well to neuroleptics. Both drugs induced significantly more parkinsonism than placebo, but differently so: chlorpromazine induced both types of parkinsonism hypo- and hyper-kinetic symptoms, whereas remoxipride induced hyperkinetic symptoms. Chlorpromazine caused more tachycardia, drowsiness, orthostatic dizziness, and dry mouth than the other two treatments, while patients on remoxipride suffered more from insomnia than those on the other two treatments.