Data are presented on the 43 people who died over a 22-year follow-up period of a cohort of 128 heroin addicts drawn in 1969 from the newly opened London clinics. The main causes of death were drug-related, with 18 deaths specifically determined as due to overdose, of which the great majority were among people being prescribed opiates at the time. The mortality rate was a mean of 1.84% annually, and the excess mortality ratio was 11.9. This excess was highest at the beginning and varied over the period of study, appearing higher at the opening of the clinics and again in the mid-1980s. No sex differences in mortality rates were demonstrated but the excess mortality was concentrated at younger ages. No prediction of the 85 survivors could be made on the basis of length of heroin use prior to study intake, nor on age at intake.