The aim of this study was to evaluate the efficacy, safety, and tolerability of sertindole in comparison with olanzapine in patients with chronic schizophrenia who did not respond successfully to their previous treatments. Patients with schizophrenia who were at least moderately ill and had failed to respond to previous antipsychotic treatment were randomized to double-blind sertindole or olanzapine treatment. A total of 389 patients were treated, 196 with sertindole (mean dose=17 mg/day) and 193 with olanzapine (mean dose=16 mg/day). Both drugs improved all the efficacy scale scores including the Positive and Negative Syndrome Scale total score. Although sertindole failed to prove noninferiority to olanzapine in terms of reduction in PANSS total score with the last-observation-carried-forward analysis, this can be attributed to the higher withdrawal rate in the sertindole group by day 16 by which sertindole was up-titrated to the effective dose. On excluding early withdrawals, the noninferiority criterion was fulfilled, as also in the observed-case analysis. They had similar safety profiles with respect to the total incidence of adverse events. The incidence of asymptomatic QT prolongation was higher in the sertindole group. Sertindole has an efficacy and safety profile that is comparable to that of olanzapine. The slow titration schedule and lack of sedating effect of sertindole should be considered when initiating treatment with this drug.