The discrepancy in HIV rates among Eastern and Central European injecting drug users (IDUs) suggests that, in addition to risk behaviors, social contact patterns also play an important role. We identify two groups of IDUs in Budapest, Hungary, marginalized IDUs (M-IDUs) and socially integrated IDUs (SI-IDUs), and compare their HIV/HBV/HCV social and risk network characteristics, risk behaviors, and travel patterns. Between 05/2003 and 01/2004, 29 non-treatmentrecruited young IDUs in Budapest participated in ethnographic interviews and focus groups. The mean age was 23.6 years (SD=3.6); eight were female and two Roma/Gypsy. Most injected heroin (n=23) and/or amphetamines (n=10) in the past 30 days. M-IDUs had no legal employment, injected heroin and sniffed glue, and stopped using drugs in treatment/prison. SI-IDUs had regular jobs or were students, injected heroin and sniffed cocaine, and stopped using drugs before exams/tests. Both M-IDUs and SI-IDUs shared injecting equipment on occasion and used condoms rarely. M-IDUs had a large social network of "buddies" and a small risk network of "friends". SI-IDUs had two separate large social networks of "buddies": a M-IDU and a non-IDU network; and a small risk network of "friends". Both groups reported monogamous sexual relationships. M-IDUs traveled within Hungary whereas SI-IDUs traveled to Western Europe. If an HIV epidemic among IDUs in Hungary is not prevented, SI-IDUs may form a potential "bridge" of HIV infection between highrisk IDU populations and the low-risk, general population, while M-IDUs may become cores of infection. Different approaches may be appropriate for M-IDUs and SI-IDUs to prevent HIV.