Racial/ethnic disparities in mental disorders, including pathological gambling disorder (PGD), may be either real or artifacts of how they are conceptualized and measured. We aimed to assess racial/ethnic variation in the reliability of self-reported lifetime PGD determined by meeting > or = 5 criteria of the Diagnostic and Statistical Manual of Mental Disorders. Using community advertising, we recruited 15-85-year-old Caucasians (n = 225) and African (American/other minorities (n = 87), who had gambled more than 5 times lifetime), for 2 interviews, held 1 week apart, about gambling and associated behaviors. Results indicate substantial to almost-perfect DSM-IV PGD reliability for Caucasians (kappa = 0.82) and African Americans/other minorities (kappa = 0.68). Reliability for symptoms and for game-specific disorders was fair to almost perfect (kappa = 0.37-0.90). After adjusting results for confounding variables and multiple comparisons, racial/ethnic variation in PGD and game-specific reliability failed to persist. Implications exist for increased attention to screening and prevention efforts critical to reducing racial/ethnic disparities in PGD prevalence.