Background. Little is known about the association of rheumatic heart disease (RHD) with incident heart failure (HF) among older adults. Design. Cardiovascular Health Study, a prospective cohort study. Methods. Of the 4,751 community-dwelling adults ≥ 65 years, free of prevalent HF at baseline, 140 had RHD, defined as self-reported physician-diagnosed RHD along with echocardiographic evidence of left-sided valvular disease. Propensity scores for RHD, estimated for each of the 4,751 participants, were used to assemble a cohort of 720, in which 124 and 596 participants with and without RHD, respectively, were balanced on 62 baseline characteristics. Results. Incident HF developed in 33% and 22% of matched participants with and without RHD, respectively, during 13 years of follow-up (hazard ratio when RHD was compared to no-RHD 1.60; 95% confidence interval 1.132.28; P = 0.008). Pre-match unadjusted, multivariable-adjusted, and propensity-adjusted hazard ratios (95% confidence intervals) for RHD-associated incident heart failure were 2.04 (1.542.71; P < 0.001), 1.32 (1.021.70; P = 0.034), and 1.55 (1.142.11; P = 0.005), respectively. RHD was not associated with all-cause mortality (HR 1.09; 95% CI 0.821.45; P = 0.568). Conclusion. RHD is an independent risk factor for incident HF among community-dwelling older adults free of HF, but has no association with mortality. © 2012 Informa UK, Ltd.