This study investigated the long-term survival rates of 85 patients with hereditary hemochromatosis. Eighty-five patients with documented hereditary hemochromatosis diagnosed between 1958 and 1989 and followed up at the University Hospital (University of Western Ontario) medical center were retrospectively reviewed for this analysis. The current status of the patient was assessed by interview or written questionnaire completed by the patient or the family physician. Estimates of differences in survival rates were obtained using Kaplan-Meier life-table and Cox regression analysis. Liver histology, clinical features of the disease, and number of venesections were analyzed to determine their relationship to survival. In the course of a mean follow-up interval of 8.1 +/- 6.8 years (range, 0-31 years), there were 17 deaths among the 85 hemochromatosis patients. Patients with cirrhosis at the time of diagnosis were 5.5 times more likely to die than noncirrhotic patients. Patients who were noncirrhotic at the time of diagnosis had an estimated survival that was not significantly different from age- and sex-matched members of the normal population. Diabetes did not increase the risk of death after data were controlled for the presence of cirrhosis. Early diagnosis and treatment of hemochromatosis in the precirrhotic stage can lead to long-term survival similar to that in the general population. The presence of cirrhosis significantly increases mortality and is the major clinical factor affecting survival.