Although public health surveillance system data are widely used to describe the epidemiology of communicable disease, occurrence of hepatitis B and C virus (HBV and HCV, respectively) infections may be misrepresented by under-reporting in injection drug users (IDUs). This study was carried out to examine the relationship between HBV and HCV incidence and case-reporting of hepatitis B and C in Seattle IDUs. Names of participants in a Seattle IDU cohort study who acquired HBV or HCV infection over a 12-month follow-up period were compared to a database of persons with acute hepatitis B and C reported to the health department surveillance unit over the same period. Of 2,208 IDUs enrolled in the cohort who completed a follow-up visit, 63/759 acquired HBV infection, 53/317 acquired HCV infection, and 3 subjects acquired both HBV and HCV. Of 113 cohort subjects who acquired HBV or HCV, only 2 (1.5%) cases were reported; both had acute hepatitis B. The upper 95% confidence limit for case-reporting of hepatitis C in the cohort was 5.7%, and for hepatitis B, it was 7.5%. In this study, reporting of acute hepatitis in IDUs was extremely low, raising questions regarding the use of community surveillance data to estimate underlying incidence in that population group.