Death is a major outcome measure, and the value of an audit of the preventable factors leading to individual deaths is accepted in secondary care. Although approximately 20 of a typical general practitioner's patients will die every year, audits of deaths in general practice are seldom published. This is probably partly due to the difficulties posed by comparability, partly due to the wide range of causes of such deaths and possible contributory factors, and partly due to confidentiality considerations. We describe an audit of all death occurring in our practices over a 12-month period. Although cigarette smoking was considered to exceed all other contributory causes of death, preventable factors occurred in both primary and secondary care sufficiently often for important conclusions to be drawn. Comparison of causes of death between general practices is possible, and such enquiry can demonstrate important quality issues for both primary and secondary care.