Background. Stroke is a disease of high prevalence, the second cause of death and the leading cause of chronic disability in adults in the world. Cardiac rehabilitation programs (CRP) with their components of aerobic exercise and education for modification of risk factors are potentially beneficial for secondary prevention in stroke patients. Objective. To estimate the aggregate effect of CRP, compared to conventional rehabilitation, on the recurrence of stroke and the modification of clinical and functional variables in adult patients diagnosed with stroke. Methods. A systematic literature search was performed. Randomized clinical trials using CRP or conventional rehabilitation, in patients older than 18 years, survivors of stroke, were included. Information of clinical and functional outcomes from primary studies was obtained, a fixed effects model was used to calculate the estimated effect for outcomes that had at least two primary studies. Results. Five randomized clinical trials were included. A total of 409 patients with average age of 65.4 years (SD = 9.1) in the intervention group and of 66 years (SD=7.5) in the control group, 60% male. The aggregate effect of a CRP compared to a control intervention (35 patients receiving CRP and 35 receiving a control intervention), suggests a non-significant reduction for the group undergoing CRP in the Framingham scale score (-3.90 points, 95% CI -7.98, 0.18), total cholesterol level (-0.20 mmol / L, 95% CI -0.48, 0.09), systolic blood pressure (-0.97 mmHg, 95% CI -9.49, 7.56) and diastolic blood pressure (1.00 mmHg, 95% CI -4.05, 6.06). Three of the five studies were rated as "low risk" of bias. Conclusion. The results consistently but inaccurately suggest a reduction in cardiovascular risk score, cholesterol level and systolic and diastolic blood pressure values in patients undergoing CRP. More robust randomized clinical trials evaluating long-term clinical outcomes are required.