Purpose: To evaluate (1) whether there were racial differences in the androgen receptor gene CAG repeat length and in clinical or laboratory attributes of prostate cancer at the time of diagnosis; (2) whether there were differences in race, Gleason score, prostate-specific antigen (PSA) level, and stage at diagnosis by androgen receptor gene CAG repeat length; and (3) whether sociodemographic, clinical, and laboratory based factors might be associated with advanced-stage prostate cancer. To our knowledge, our study is the first to report on CAG repeat lengths in a cohort of prostate cancer patients, which includes large numbers of African-American men. Methods: CAG repeat lengths on the androgen receptor gene were evaluated for 151 African-American and 168 white veterans with prostate cancer. The X2 test, t test, and logistic regression analyses were used to evaluate the associations between CAG repeat lengths and race, stage, histologic grade, and PSA levels at diagnosis. Results: The mean age of the cohort at the time of diagnosis was 68.7 years. At presentation, 42.0% had stage D prostate cancer, 26.5% had Gleason scores of 8 to 10, and 53.0% had PSA levels ≥ 10 ng/dL. Mean androgen receptor gene CAG repeat length for white veterans was 21.9 (SD, 3.5) versus 19.8 (SD, 3.2) for African-American veterans (P = .001). Men with shorter CAG repeats were more likely to have stage D prostate cancer (P = .09) but were not more likely to have a higher PSA concentration or Gleason score. Conclusion: In this cohort of men with prostate cancer, short CAG repeat length on the androgen receptor gene was associated with African-American race and possibly with higher stage but not with other clinical or pathologic findings.