One hundred and seventy-two patients with partial or complete hypopituitarism diagnosed between 1967 and 1994 were studied retrospectively. Those with acromegaly or Cushing's disease were excluded. One hundred and two patients were male (median age at, diagnosis, 53 yr; range, 12-78 yr) and 70 female (median age at diagnosis, 51 yr; range, 1-74 yr). In 131 patients the cause of hypopituitarism was a pituitary tumor or the effects of its treatment, as the majority underwent surgery and/or radiotherapy. In 22, the cause was an extrapituitary tumor, 14 were termed idiopathic, 2 developed hypopituitarism as a result of basal sarcoid, 2 were due to trauma, and 1 was the result of Sheehan's syndrome. The patients were treated with standard replacement therapy. Mortality due to all causes was higher than expected in an age- and sex-matched control population (ratio of observed/expected deaths, 1.73; 95% confidence interval, 1.28-2.28; P < 0.01). Females tended to have a worse prognosis (ratio of observed/expected deaths, 2.29; 95% confidence interval, 1.37-3.58; P < 0.01) than their male counterparts (ratio of observed/expected deaths, 1.50; 95% confidence interval, 1.02-2.13; P < 0.01). There was a small but nonsignificant increase in the number of deaths due to vascular disease (ratio of observed/expected deaths, 1.35; 95% confidence interval, 0.84-2.07; P = 0.11). The only significant independent predictive factors for survival were age at diagnosis and hypogonadism. The majority of the male hypogonadal cohort received replacement therapy (79%), but fewer of the females did so (27%). Hypogonadal patients had a better prognosis than their eugonadal counterparts (log rank, 6.85; P < 0.01). Our data confirmed that mortality in patients with hypopituitarism is significantly increased. However, the contribution of vascular disease to this poor prognosis was not as great as previously reported, and overall, our results favor a multifactorial explanation of the poor long term outcome.