AIMS--Cataract surgery has at times been said to correlate with an increased death risk. We have therefore analysed the standardised death ratio in a population based cohort of patients that had undergone cataract surgery. METHODS--Data for all patients undergoing cataract surgery from 1986 up to and including 1990 in the Lund Health Care District were prospectively recorded, and 5120 were retrieved for analysis. Death dates and primary death diagnoses for each patient were obtained from the Swedish Bureau of Census up to and including 1991. Standardised mortality ratios were calculated for all patients, subdivided into a number of categories: inpatients, outpatients, sex, age or = 75 years, patients with heart and circulatory diseases, with malignant tumours, and with diabetes. The cut off age was set at 74 because this divided the patients into two approximately equal groups. Using time dependent survival regression, the relative risk for dying were estimated for sex, age, and for postoperative YAG laser capsulotomy, and also for diabetic patients and patients with rheumatoid arthritis. RESULTS--Inpatients almost always show an increased standardised mortality ratio compared with outpatients. Young patients and diabetic patients also showed an increased standardised mortality ratio, compared with the normal population, but not older patients, who constitute the majority. Cardiovascular death diagnoses were overrepresented among the young. CONCLUSIONS--Cataract surgery is correlated with an increased standardised mortality ratio only in young patients and in patients with certain complicating diseases like diabetes and cardiovascular diseases.