The calculation and comparison of relative survival rates after interventional studies is a method that permits correction for important demographic variables, thereby adjusting for the "background mortality" in the general population. Long-term relative survival rates were analyzed in a consecutive series of 2,805 Swedish patients who, on the basis of clinical symptoms, underwent aortic valve replacement (n = 1,741), mitral valve replacement (n = 792) and double (aortic plus mitral) valve replacement (n = 272) between 1969 and 1983. The follow-up period, which closed August 1, 1985, included 100% of patients and covered 16,822 patient-years. Autopsy was performed in 75% of all deaths. The results underscore previously well known differences between the long-term survival after aortic valve replacement and mitral or double valve replacement, whereas no differences were noted between mitral and double valve replacement. Within the subgroup undergoing aortic valve replacement, analysis of relative survival rates disclosed a highly significant (p less than 0.001) difference between patients operated on for aortic stenosis and those operated on for aortic regurgitation, representing a mortality rate more than twice as high in the latter group. This difference was of much lesser magnitude when analyzed in the standard (actuarial) way. With a low (less than 2.5%) operative mortality rate for patients undergoing isolated elective aortic valve replacement in the current era and with an acceptable incidence of late valve-related death (5.2% at 10 years), these results may justify aortic valve replacement earlier in the course of chronic aortic regurgitation to prevent irreversible myocardial damage.(ABSTRACT TRUNCATED AT 250 WORDS)