BACKGROUND: A regulatory signal suggested that sertindole was associated with a 10-fold increased risk of death. This observational study aimed to identify deaths and their causes in patients treated with sertindole. METHODS: Sertindole-treated patients in Germany, Austria, Belgium, Hungary, The Netherlands and the United Kingdom were identified and their vital status ascertained. Eventual cause of death and patient and treatment characteristics were ascertained. A nested case-control study compared on-treatment cardiac or unexplained deaths with randomly selected survivors matched on age and gender for potential risk factors. RESULTS: Eight thousand six hundred eight patients were identified as having initiated treatment with sertindole, for a total 3819 person-years exposure. Thirty-five had died since (all-cause mortality rate 0.92 per 100 person-years exposed). Of these, 15 (43%) had been previously reported to the regulatory authorities. Eight deaths were suicides, of which 2 (25%) had been previously reported; 11 were cardiac deaths, of which 8 (73%) had been reported. On average, patients who died because of cardiac disorders were older, and patients who committed suicide were younger than the other patients who died. The nested case-control study revealed that patients receiving sertindole were at a higher risk of premature cardiac or unexplained death if they had hypertension or other cardiovascular disorders associated with diabetes or metabolic disorder. CONCLUSION: Overall and cardiovascular death rates were similar or lower than those observed in premarketing clinical trials. There was a very high reporting rate of all-cause and cardiovascular deaths that could have contributed to the regulatory alert.