OBJECTIVES: Prostate cancer (PCa) screening has generated considerable controversy after recent recommendations to discontinue the use of the prostate-specific antigen (PSA). Our objective was to propose effective screening strategies for promoting early detection of PCa. METHODS: We applied generalized odds rate hazards (GORH) model for interval-censored data to the PCa subset of the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening trial data. The final regression analysis on the PLCO data provided accurate estimates of the effects of significant risk factors for PCa and an estimate of the cumulative incidence function (CIF) which was used to propose the screening strategy. Participants with <2 PSA tests and those with missing values in important risk factors were excluded. Risk models depicting probabilities of developing PCa in the next 't' years were generated using 5 year age increments. RESULTS: The final dataset consisted of 33735 subjects. Majority of participants were on average 62 years of age, non-hispanic White males. About 89% of participants had not developed PCa during the course of the trial. Most participants (55%) had undergone 6 PSA tests. About 8% had undergone at least 1 biopsy. Age, race, baseline PSA, biopsy, diabetes and family history of PCa were identified as significant risk factors for the model used to develop the screening strategy. For men studied in 5 year increments beginning at age 50 with a family history of PCa and concurrent diabetes, Asian men were at the lowest risk with significant increase in risk occurring between 6-7 years of follow up, followed by White men. African American men had the highest risk of developing PCa. CONCLUSIONS: Our findings have significant implications for risk assessment for PCa by race and time elapsed since 50. How the risk factors and PSA values affect likelihood of PCa can help identification of specific and sensitive screening strategy.