Comparisons of patient and technique survival were made for 120 CAPD and 139 HD patients undergoing dialysis between January 1981 and December 1986. Cox's proportional hazard regression model was used to compare patient and technique survival, with an adjustment for pre-treatment prognostic differences. Only the patients' first treatments were considered. The CAPD patients were 10 years older, on the average, than the HD patients and had more complicated conditions (58% with 3 or more co-existing risk factors vs. 35%). Overall patient survival between CAPD and HD did not differ (P = 0.2694). However, when adjusted for patient age, sex and other comorbid complicating conditions, CAPD patients over the age of 66 had a significantly lower risk of death than their HD counterparts (P less than 0.05). There were no differences in the adjusted patient survival for patients aged 30 to 66. Four pre-treatment prognostic factors had statistically significant adverse effects on patient survival: age, diabetes, malignancy and peripheral vascular disease. Survival of the HD technique, when unadjusted, was better than survival of CAPD (P = 0.0457). Even after adjustment for sex and age, this difference was still very nearly significant (P = 0.0656). No risk factors were found to be significantly associated with technique survival. Based on patient and technique survival, CAPD would appear to be an excellent alternative to HD and may be the preferred treatment for high risk patients over the age of 66.