The primary objective of this project is to develop an integrated statistical approach to modeling, estimation and prediction of the therapeutic benefit and population impact of prostate cancer surveillance. The approach we propose allows us to estimate contributions of different effects ofsurveillance on diagnosis and survival of patients with prostate cancer and population processes. These effects are the curative effect of surveillance, lead time effect and over-diagnosis. We will evaluate the effect of heterogeneity and associated over-optimistic selection bias on the observedprostate cancer incidence and mortality. Quantitative interpretation for the observed incidence and mortality trends in prostate cancer will be provided. Predictions of the impact of intensifiedsurveillance and changes in population heterogeneity on prostate cancer incidence, mortality, andthe time natural history of the disease will be made. The approach will be applied to data onprostate cancer from the Utah Population Database, the Utah Cancer Registry, public data from theSEER program, and clinical data from the Memorial Sloan Kettering Cancer Center. Simulations using mechanistic models of prostate cancer and screening will be used to validate our models.