In order to determine whether acromegaly is still associated with increased mortality, a hospital case note review of all patients with acromegaly followed up in Stoke-on-Trent since 1967 was carried out. Of 79 subjects identified, 51 are alive and being monitored and 28 have died. Mortality was compared to the general population by life table analysis. Secretion of growth hormone was assessed and compared in dead and alive patients. The effect of diabetes, hypertension, and growth hormone secretion on long-term outcome was assessed. Acromegaly is still associated with increased mortality, with an overall ratio of observed to expected deaths equal to 2.68 (95% C.I. 1.8-3.9; p < 0.001), but the survival of 31 (39%) patients whose growth hormone level had been reduced to below 5 mU/l was equal to that of the general population (O/E = 1.42; 95% C.I. 0.46-3.31: p > 0.05). The dead patients had had significantly higher growth hormone levels than those still alive, but mortality did not appear to be influenced by diabetes or hypertension. The cause of death was vascular in 57% of cases. Growth hormone hypersecretion is still associated with excess mortality in acromegaly. The present study suggests that the therapeutic objective should be to lower average daytime growth hormone levels to less than 5 mU/l. There is need for a large study to compare different modes of treatment in terms of their effect on growth hormone secretion and on long-term outcome.