AbstractThere is an increasing debate regarding the frequency of prostate‐specific antigen (PSA) testing and the current primary screening modality used to detect prostate cancer. The purpose of this study is to determine whether PSA screening intervals should be based on initial PSA. Our study explores longitudinal changes of PSA levels in black and white males separately. Study participants were 768 white and 450 black males attending an annual prostate cancer screening. We fit a longitudinal repeated measures model separately for blacks and whites and estimated the probability of PSA converting to greater than 4.0 ng/ml at a follow‐up year given baseline PSA range among males without an abnormal DRE. Black and white males with a baseline PSA between 0 and 1.0 ng/ml, with a healthy or enlarged prostate, have a less than 1% chance that their PSA will increase above 4.0 ng/ml over the following 5 years. Black and white males with a PSA between 1.0 and 1.9 ng/ml have a less than 1% chance of PSA conversion to greater than 4.0 ng/ml over the next year. Our findings further support that annual screening for prostate cancer may not be necessary, specifically males with a baseline PSA less than 2.0 may not need to undergo annual screening. Our results suggest that race does not affect the longitudinal trend of PSA enough to warrant setting screening intervals based on race. © 2005 Wiley‐Liss, Inc.