The purpose of this study was to investigate the prevalence of gambling disorder and comorbid psychiatric disorders in a homeless population and identify features related to potential subtypes. At baseline, participants were administered a structured interview including socio-demographic sections of the National Comorbidity Study (NCS) interview; seven diagnostic sections of the Diagnostic Interview Schedule (DIS); the alcohol and drug abuse sections of the Composite International Diagnostic Interview-Substance Abuse Module (CIDI-SAM); and the Homeless Supplement to the DIS. At nine months post-baseline assessment, participants were administered additional NCS family history questions and the South Oaks Gambling Screen (SOGS). Participants were an epidemiologic sample of 275 predominately African-American homeless individuals, grouped as lifetime non-gamblers (n = 60), recreational gamblers (n = 152), and problem gamblers (n = 63), recruited on the street and through homeless shelters. Results indicate that lifetime rates of sub-clinical problem (46.2%) and disordered (12.0%) gambling were significantly higher than in the general population. Problem gamblers were more likely than non-problem gamblers to meet diagnostic criteria for antisocial personality disorder, post-traumatic stress disorder, bipolar disorder, and any psychiatric disorder, and more likely than non-gamblers to use illicit drugs or meet criteria for abuse/dependence for nicotine, alcohol, or any substance. This study provides evidence that problem gambling is a significant public health issue among the African-American homeless population. Homeless services should include assessment for problem gambling along with psychiatric disorders and referrals to resources and treatment programs. Future studies should explore the relationship of the onset and course of problem gambling and other psychiatric disorders with homelessness as well as racial differences in gambling patterns and problem severity over time.