A case-control study was conducted to investigate the association between physical activity and prostate cancer risk in China. A total of 130 recent histopathologically confirmed cancer cases and 274 controls were interviewed. Information collected included type, intensity and duration of various physical activities. The adjusted odds ratios of prostate cancer risk for physical activity variables were obtained from multivariate logistic regression models. Results showed that moderate physical activity was inversely related to the prostate cancer risk, with adjusted odds ratio being 0.20 (95% confidence interval 0.07-0.62) for the upper vs. the lower quartiles of weekly metabolic equivalent task-hours. The dose-response relationship was also significant (p = 0.015). The findings suggest that moderate physical activity may be protective against prostate cancer for Chinese men.