BACKGROUND Levels of physical fitness are low after stroke. It is unknown whether improving physical fitness after stroke reduces disability. OBJECTIVE To determine whether fitness training after stroke reduces death, dependence, and disability. The secondary aims were to determine the effects of training on physical fitness, mobility, physical function, quality of life, mood, and incidence of adverse events. METHODS We searched the Cochrane Stroke Group Trials Register (last searched April 2010), the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, July 2010), MEDLINE (1966 to March 2010), EMBASE (1980 to March 2010), CINAHL (1982 to March 2010), SPORTDiscus (1949 to March 2010), and five additional databases (March 2010). We also searched ongoing trials registers, handsearched relevant journals and conference proceedings, screened reference lists, and contacted experts in the field. METHODS Randomised trials comparing either cardiorespiratory training or resistance training, or both, with no intervention, a non-exercise intervention, or usual care in stroke survivors. METHODS Two review authors independently selected trials, assessed quality, and extracted data. We analysed data using random-effects meta-analyses. Diverse outcome measures limited the intended analyses. RESULTS We included 32 trials, involving 1414 participants, which comprised cardiorespiratory (14 trials, 651 participants), resistance (seven trials, 246 participants), and mixed training interventions (11 trials, 517 participants). Five deaths were reported at the end of the intervention and nine at the end of follow-up. No dependence data were reported. Diverse outcome measures made data pooling difficult. The majority of the estimates of effect were not significant. Cardiorespiratory training involving walking improved maximum walking speed (mean difference (MD) 8.66 metres per minute, 95% confidence interval (CI) 2.98 to 14.34), preferred gait speed (MD 4.68 metres per minute, 95% CI 1.40 to 7.96) and walking capacity (MD 47.13 metres per six minutes, 95% CI 19.39 to 74.88) at the end of the intervention. These training effects were retained at the end of follow-up. Mixed training, involving walking, increased preferred walking speed (MD 2.93 metres per minute, 95% CI 0.02 to 5.84) and walking capacity (MD 30.59 metres per six minutes, 95% CI 8.90 to 52.28) but effects were smaller and there was heterogeneity amongst the trial results. There were insufficient data to assess the effects of resistance training. The variability in the quality of included trials hampered the reliability and generalizability of the observed results. CONCLUSIONS The effects of training on death, dependence, and disability after stroke are unclear. There is sufficient evidence to incorporate cardiorespiratory training involving walking within post-stroke rehabilitation programmes to improve speed, tolerance, and independence during walking. Further well-designed trials are needed to determine the optimal exercise prescription and identify long-term benefits.