The objective of the European Post-marketing Observational Serdolect((R)) (EPOS) Study was to compare the safety of treatment with Serdolect (sertindole) with that of usual treatment in patients with schizophrenia, in normal European clinical practice. The EPOS was a multicentre, multinational, referenced, cohort study. Patients were enrolled at 226 centres in ten European countries. The study was prematurely terminated in 1998 as a result of the temporary market suspension of sertindole. Termination of the study reduced the number of patients recruited from the planned 12,000 to 2,321. While the power of the study was weakened, it did provide useful mortality information, which may be useful for future long-term studies. Crude mortality in the sertindole and non-sertindole groups was 1.45 (95% confidence interval, CI 0.53-3.16) and 1.50 (CI 0.72-2.76) deaths/100 patient-years exposed, respectively. There were no more cardiac deaths in the sertindole group than in the non-sertindole group. QT interval prolongation did not translate into an increased risk of death. Sertindole was well tolerated and caused few extrapyramidal symptoms. Although CIs remained large, this post-marketing study does not provide any evidence against the use of sertindole under normal conditions. Sertindole was well tolerated and posed no significant safety problems.