OBJECTIVE: To describe the epidemiology of infection with hepatitis C virus (HCV) among injecting drug users (IDUs) in Victoria. DESIGN AND SUBJECTS: Subjects were current IDUs from a wide spectrum of age, sex and social background, enrolled in a prospective study of injecting drug use. They were contacted by peer workers through their social networks and through community agencies and prisons, and were regularly followed for interview and blood collection in the field. Sera were tested for presence of antibody to HCV (anti-HCV), for the presence of HCV RNA directly in serum, and for measures of liver function. The results were correlated with demographic variables. SETTING: Rural and metropolitan Victoria. MAIN OUTCOME MEASURES: Presence of anti-HCV and demonstration of HCV RNA. RESULTS: Two-thirds (68%, 206/303) of the current cohort of IDUs were seropositive for HCV, risk being particularly associated with duration of injecting, and independently for men with opiate use and prison history, and for women with a history of methadone therapy. HCV RNA was detected in 48% (76/160) by polymerase chain reaction (PCR); 61% (74/122) of these subjects were HCV seropositive and 5% (2/38) seronegative. Of 32 HCV seronegative subjects followed for a mean period of 291 days, five seroconverted to HCV, an incidence of 20 infections per 100 person-years. Those who seroconverted were older, more likely to be male, had been injecting longer, more often reported opiate use, and were more likely to be based in the country. Serum liver enzyme levels were higher and more likely to be abnormal in HCV seropositive than seronegative subjects, and were highest in those seropositive subjects in whom HCV RNA was detected. CONCLUSIONS: This population of IDUs has a very high rate of exposure to HCV, related to duration of injecting and independently to opiate use and prison history, perhaps reflecting increased risk in particular social networks. There is evidence of high rates of carriage of HCV, of continuing transmission of HCV, and of ongoing liver disease among these IDUs. If these IDUs are at all representative of all IDUs in Australia, we estimate that 80,000 current and former IDUs may be at risk of chronic liver disease from HCV, and that 8000-10,000 new infections may be occurring each year. Two subjects who were seronegative had HCV RNA detectable in sera. These data have important implications for screening programs and document the need for further measures to prevent spread of blood-borne viruses including HIV among IDUs.