OBJECTIVES:  The benefits of exercise for cardiovascular health are widely known. A recent meta-analysis reported an association between exercise and a reduction in the risk of cancer. This study aims to assess the cost-utility of exercise for the prevention of cancer from a US health-system perspective. METHODS:  A cohort life-table analysis was developed to model life years (LYs) and quality-adjusted life years (QALYs) of low versus high physical activity over a lifetime horizon in US adults. Hazard ratios for various cancers (bladder, breast, colorectal, endometrial, esophageal, gastric, head/neck, kidney, liver, lung, leukemia and myeloma) were obtained from a meta-analysis. Age- and sex-specific incidence and mortality rates were obtained from SEER. Utility weights, baseline healthcare costs, and attributable cancer costs were obtained from the literature. Three exercise programs of varying costs were considered including exercising at home, gym membership, and gym with personal trainer. Incremental analyses, stratified by sex, and probabilistic sensitivity analyses (PSAs) were conducted. RESULTS:  High activity levels, compared to low, increased undiscounted LYs by 0.37 and 0.34 years and discounted QALYs by 0.07 and 0.06 years for males and females, respectively. High activity reduced lifetime cancer costs by $1,873 and $1,593 and lifetime healthcare costs by $779 and $462, respectively, for males and females. The incremental cost per QALY inclusive of low, moderate and high exercise costs were $88,780, $279,980, and $825,071 for males and $99,063, $300,893 and $876,288 for females, respectively. In the PSA low cost exercise had a 15.6% and 11.9% probability of being cost-effective for males and females, respectively, at a $50,000/QALY threshold. CONCLUSIONS:  Despite many associated health benefits, exercise may not be a cost-effective means for cancer prevention – this may have implications for potential pharmacologic interventions to prevent cancer. Given the limitations of observational study data, further research is warranted to confirm these findings.