OBJECTIVE To update mortality rates and long-term survival of patients admitted to the hospital with ruptured abdominal aortic aneurysm (AAA) and to study prognostic factors associated with mortality. METHODS Retrospective follow-up. METHODS 309 patients (274 men, 35 women, average age 71) admitted to the hospital between January 1980 and January 1994 who were surgically treated for ruptured AAA were studied. METHODS To identify the preoperative (9), intraoperative (23) and postoperative (49) variables associated with mortality logistic regression analysis (mortality within 48 h) and Cox regression analysis (mortality between 48 h and 30 days) were performed. RESULTS Hospital mortality improved from 1980 to 1994. Compared with the normal population adjusted for age and sex the long-term mortality rate was increased (standardised mortality ratio 2.1; 95% confidence interval 1.7-2.5). Increased age, peroperative hypotension and need for a bifurcated graft were associated with significantly increased mortality. Co-morbidity was not a predictive variable. Overall hospital mortality was 25%. CONCLUSIONS Surgical repair of ruptured AAA should be considered even in patients with co-morbidity. Elderly patients with severe preoperative hypotension have a very high mortality rate and surgery may not be justified in these cases. Long-term survival is also worse in older patients.