BACKGROUND: The purpose of this meta-analysis was to explore the efficacy of exercise as a nonpharmacologic intervention to reduce cancer-related fatigue (CRF) among adult cancer survivors. We also investigated how different components of the exercise prescription (Ex R(x)), methodologic considerations, and subject characteristics modulate CRF. METHODS: A systematic search for randomized controlled trials was conducted using words related to cancer, exercise, and fatigue. RESULTS: In total, 44 studies with 48 interventions qualified, including 3,254 participants of varying cancer types, stages of diagnosis, treatments, and exercise interventions. Cancer survivors in exercise interventions reduced their CRF levels to a greater extent than usual care controls, d(+) = 0.31 (95% CI = 0.22-0.40), an effect that appeared to generalize across several types of cancer. CRF levels improved in direct proportion to the intensity of resistance exercise (β = 0.60, P = 0.01), a pattern that was stronger in higher quality studies (β = 0.23, P < 0.05). CRF levels also reduced to a greater extent when interventions were theoretically driven (β = 0.48, P < 0.001) or cancer survivors were older (β = 0.24, P = 0.04). CONCLUSIONS: Exercise reduced CRF especially in programs that involved moderate-intensity, resistance exercise among older cancer survivors and that were guided by theory. IMPACT: Our results indicate exercise interventions for adult cancer survivors should be multi-dimensional and individualized according to health outcome and cancer type.