Physically active individuals have lower rates of many cancers and improved cancer outcomes. Controlled exercise trials measuring putative biomarkers of cancer risk are being conducted to further understand the role of exercise in cancer etiology and progression. We aimed to systematically review the effect of exercise on various biomarkers. A comprehensive search strategy identified 353 publications from January 1980 to August 2010. We included those clinical trials of exercise measuring biomarkers following minimum 4-week intervention among cancer survivors or people with one or more cancer risk factors. Two reviewers abstracted data and assessed quality independently. Effect sizes and 95% confidence intervals were estimated. Four primary prevention and five tertiary prevention trials were included. Exercise had a small to moderate effect on improving concentrations of several blood biomarkers implicated in breast and colon cancer pathways including insulin, leptin, estrogens, and apoptosis regulation. In breast cancer survivors, exercise had a small to moderate effect on improving some biomarkers associated with prognosis including various insulin-like growth factor axis proteins, insulin, and inflammation; and a large effect on enhancing immune function. Data are few, but there is some evidence to support the role of exercise in modulating various cancer pathways.