OBJECTIVES: Exercise training is an effective strategy to reduce combined clinical outcomes in heart failure (HF). Nonetheless, implementation of such programs has been restricted to university and specialized centers. Economic analysis of this intervention may help reinforce health policy worldwide. METHODS: Using a Markov model, we performed a cost-effectiveness analysis to estimate the costs, health gains, and incremental cost-effectiveness (international dollars [I$] per quality-adjusted life year [QALY] gained) of an exercise training program in HF class II and III NYHA patients, comparing with standard treatment, assuming a public system perspective in Brazil. QALYs were estimated from an outpatient cohort of 318 patients. Treatment efficacy was obtained from controlled trials and meta-analysis; treatment costs were derived from published data and National Health System reimbursement rates in 2008. Exercise training costs were obtained from a cardiac rehabilitation center. Robusteness of results was tested by Monte Carlo simulation and sensitivity analysis. RESULTS: Considering a 35% reduction of mortality with exercise training and an annual cost of I$ 1,176 per patient, this strategy had a total cumulative cost of I$ 25,856 and 4.95 QALYs. Comparing with standard treatment, which had a total cost of I$ 16,758 and 4.34 QALYs, the incremental cost per QALY of exercise training was I$ 14,965. Results were sensitive to intervention-related costs and effect size. Considering the results of the HF-ACTION trial, in a time-limited exercise program with a 11% combined event (all-cause mortality or hospitalization) reduction and an exercise cost of I$ 470 per patient, incremental cost-effectiveness ratio would be I$ 19,828/QALY. CONCLUSIONS: Under several assumptions, exercise training appeared to be cost-effective, and to offer good value for money compared to other well-accepted HF treatment strategies. The results support implementing such intervention as part of public health efforts to improve HF management.