In an earlier paper (Fichter, M. M., Quadflieg, N. (1999). Alcoholism in homeless men in the mid-nineties: results from the Bavarian Public Health Study on Homelessness. European Archives of Psychiatry and Clinical Neuroscience 249:34-44), we reported data on alcoholism and comorbidity in 265 homeless men in Munich. There-as in this paper-we divided the sample into three groups based on a lifetime diagnosis of alcohol dependence (N = 187), alcohol abuse (N = 17), and no diagnosis of alcoholism (N = 61) at baseline assessment. This study reports a three-year prospective longitudinal assessment of the original representative sample of homeless men in Munich. Interviews at baseline and at follow-up included the SCID-I and covered several other areas (cognitive impairment, somatic complaints, use of medical services, and other psychosocial variables). Of 247 homeless men still alive, at three-year follow-up, 185 (74.9%) were successfully traced and personally interviewed. Alcohol dependency in homeless men at first wave assessment (as compared to men not manifesting alcohol abuse or dependence) was associated with a higher proportion of homelessness at three-year follow-up, an increase of alcohol consumption at three-year follow-up, reduction of monthly income, higher death rate, and high use of general medical services but very low utilization of (specific substance) user treatment services. Alcoholism in homeless men constitutes a posited risk factor for an unfavorable course over time with regard to such a person's living situation and health status in spite of more utilization of medical services.