Combined data on efficacy were available from 12 double-blind short-term (maximum 8 weeks) trials comparing risperidone and other antipsychotics in patients with chronic schizophrenia. Patients received risperidone (n = 1056) or other antipsychotics (n = 703). Haloperidol (n = 473) was the most frequently prescribed other antipsychotic. Efficacy assessments include the Positive and Negative Syndrome Scale (PANSS) total, subscale (positive symptoms, negative symptoms and general psychopathology), cluster (cognitive and affective symptoms) and item (anxiety and hostility) scores. At endpoint, the mean decrease from baseline in PANSS total scores was significantly greater for patients receiving risperidone (-20.9) than other antipsychotics (-16.2; P < 0.001), or the subset receiving haloperidol (-14.3; P < 0.001). Risperidone-treated patients showed a significantly greater decrease in the positive (P < 0.01), negative (P < 0.05) and general psychopathology (P < 0.001) scores than patients receiving other antipsychotics or haloperidol. Scores for cognition, affective symptoms, anxiety and hostility each improved significantly (P < 0.05) more for patients receiving risperidone than those receiving other antipsychotics or haloperidol. Efficacy data on patients with an acute exacerbation were available from seven trials (risperidone n = 372, other antipsychotics n = 285, including haloperidol n = 120). At endpoint, the mean decrease from baseline in PANSS total scores was significantly greater for patients receiving risperidone (-24.7) than other antipsychotics (-19.8, P < 0.01) including haloperidol (-19.8, P < 0.05). Risperidone-treated patients also showed a greater decrease in positive symptom scores (-7.8) than those receiving other antipsychotics (-6.3; P < 0.01) or haloperidol (-7.1). A > or = 20% reduction in PANSS total score with risperidone, haloperidol and other antipsychotics was achieved by 65.9%, 54.3% and 54.9%, respectively; a > or = 30% PANSS reduction by 54.0%, 46.6% and 46.5% of patients, respectively; and a > or = 40% reduction by 43.8%, 33.7% and 34.4% of patients, respectively. These findings are consistent with earlier findings that show risperidone is more efficacious than haloperidol for reducing the symptoms of schizophrenia.