Objective: To assess the associations of aerobic, resistance, and combined exercise with changes in insulin resistance, fasting glucose, and fasting insulin in children and adolescents who are overweight or obese. Data searches: MEDLINE via Pubmed, Cochrane-CENTRAL, SPORTDiscus, and LILACS. Study selection: Randomized clinical trials of at least six weeks of duration that evaluated the ability of exercise training to lower at least one of the following outcomes: insulin resistance-HOMA, fasting glucose, and fasting insulin in children and/or adolescents classified as obese or overweight. Data extraction and analysis: Two independent reviewers extracted data and assessed the quality of the included studies. Differences (exercise training group minus control group) in the outcomes evaluated were analyzed using a random effects model. Results: Of 1853 articles retrieved, 17 studies were included. The meta-analysis showed that physical training in general was not associated with a reduction in fasting glucose levels compared to the control, but it was associated with reductions in fasting insulin levels (-3.37μU/ml; CI 95%, -5.16μU/ml to -1.57μU/ml; I2, 54%, p=0.003) and HOMA (-0.61; CI 95%, -1.19 to -0.02; I2, 49%, p=0.040). In addition, each modality (aerobic, resistance, and combined) was compared to the control group. Aerobic exercise was associated with declines in fasting insulin levels (-4.52μU/ml; CI 95%, -7.40 to -1.65; I2, 65%, p=0.002) and in HOMA (-1.33; 95% confidence interval, -2.47 to -0.18; I2, 73%, p=0.005). Conclusions: Exercise training, especially aerobic training, is associated with the reduction of fasting insulin levels and HOMA in children and adolescents with obesity and overweight, and may prevent metabolic syndrome and type 2 diabetes.