Background. Cognitive benefits obtained from exercise in healthy populations support the idea that aerobic and resistance training (AT+RT) would confer benefit for poststroke recovery. However, there is little evidence regarding the effectiveness of such programs. Objective. To evaluate the effects of a 6-month exercise program of AT+RT on cognition in consecutively enrolled patients with motor impairments ≥10 weeks poststroke. Methods. Outcomes were measured before and after 6 months of AT+RT on 41 patients. Cognition was measured by the Montreal Cognitive Assessment (MoCA). Secondary measures included evaluation of gas exchange anaerobic threshold (ATge), body composition by dual energy X-ray absorptiometry, and depressive symptoms by questionnaire. Results. There were significant improvements in overall MoCA scores (22.5 ± 4.5 to 24.0 ± 3.9, P < .001) as well as in the subdomains of attention/concentration (4.7 ± 1.7 to 5.2 ± 1.3, P = .03) and visuospatial/executive function (3.4 ± 1.1 to 3.9 ± 1.1, P = .002). There was a significant reduction in the proportion of patients meeting the threshold criteria for mild cognitive impairment (MCI) at baseline compared with posttraining (65.9% vs 36.6%, P < .001). In a linear regression model, there was a positive association between change in cognitive function and change in fat-free mass of the nonaffected limbs (β = .002; P = .005) and change in attention/concentration and change in ATge (β = .383; P ≤ .001), independent of age, sex, time from stroke, and change in fat mass and depression score. Conclusion. A combined training model (AT+RT) resulted in improvements in cognitive function and a reduction in the proportion of patients meeting the threshold criteria for MCI. Change in cognition was positively associated with change in fat-free mass and ATge.