Importance: Limited data exist on the prevalence and prognostic importance of right ventricular (RV) dysfunction for heart failure (HF) in the general population. Objective: To assess the prevalence of RV dysfunction and its association with HF and mortality in a community-based elderly cohort. Design, Setting, and Participants: Cross-sectional and time-to-event analysis of participants in the Atherosclerosis Risks in the Community (ARIC), a multicenter, population-based cohort study at the fifth study visit from 2011 to 2013, with a median follow-up of 4.1 years. This study included 1004 elderly participants in the ARIC study attending the fifth study visit who underwent both 3-dimensional and 2-dimensional RV echocardiography. Three-dimensional echocardiography data were analyzed between September 15, 2015, and July 24, 2016. Exposures: Right ventricular ejection fraction (RVEF), RV-pulmonary artery (PA) coupling defined by the RVEF/PA systolic pressure (PASP) ratio, and RV longitudinal strain by 3-dimensional echocardiography. Main Outcomes and Measures: For cross-sectional analysis, the prevalence of RV dysfunction across ACCF/AHA HF stages (0; A, at elevated risk for HF but without structural heart disease or clinical HF; B, structural heart disease but without clinical HF; and C, prevalent HF). For time-to-event analysis, a composite of incident HF hospitalization or all-cause death among participants free of HF at visit 5. Results: Of the 1004 participants, mean (SD) age was 76 (5) years, 385 were men (38%), and 121 were black (12%). Mean (SD) RVEF was 53% (8%). Right ventricular EF, RVEF/PASP, and RV longitudinal strain were each progressively lower across advancing HF stages. Using reference limits from stage 0 participants, RVEF was abnormal in 103 asymptomatic persons with stage A HF (15%) and 27 with stage B HF (24%). Among participants free of HF at baseline, lower RVEF and worse RV-PA coupling (ie, lower RVEF/PASP ratio) both were associated with incident HF or death independent of LVEF and N-terminal pro b-type natriuretic peptide (hazard ratio, 1.20; 95% CI, 1.02-1.42 per 5% decrease in RVEF; P =.03; hazard ratio, 1.65, 95% CI, 1.15-2.37 per 0.5 unit decrease in RVEF/PASP ratio; P =.007). Conclusions and Relevance: Right ventricular function and RV-PA coupling declined progressively across American College of Cardiology Foundation/American Heart Association HF stages. Among persons free of HF, lower RVEF was associated with incident HF or death independent of LVEF or N-terminal pro b-type natriuretic peptide.