All employees of a hospital-based clinical laboratory were studied over a 26-month period for the presence of hepatitis B antigen and antibody (HBsAg and HBsAb, respectively). Each subject also had liver function tests, including measurement of serum bile acids. Because of employee turnover, 111 employees were studied only once, while the remaining 199 were studied two or more times. The average prevalence of HBsAg was 1.7% and that of HBsAb, 23.9%, and the hepatitis attack rate was about 3% per year. The prevalence of hepatitis B markers was increased with greater exposure to body fluids and tissues and with increasing duration of employment (16%<5 years, 34%>5 years), while sex differences were not significant. Prior transfusion increased combined HBs marker prevalence from 19.4% to 39%, while a history of hepatitis within eight years increased this figure to 93%. In those who had had hepatitis more than eight years previously, HB marker prevalence was only 33%. HBsAb titers <5,000 (Abbott Laboratories, AUSAB RIA) led to loss of Ab within a year in 15/16 persons, while when titers were >5,000 Ab persisted in 26 of 34 persons. Because of the loss of Ab, spot prevalence figures underestimate the incidence of hepatitis B virus infection. The clinical significance of a positive HB marker with or without other risk factors (e.g., alcohol ingestion, oral contraceptives) was negligible, since in this whole laboratory population major abnormalities disclosed by liver function tests were rare except during acute hepatitis, e.g., serum glutamic oxaloacetic transaminase (SGOT) exceeded 60 (none>100) IU/dl in only 3/310, alkaline phosphatase>150 IU/dl in only one, and serum bile acids >15 μmol/l in 3/310. The incidence of abnormal liver function test results was not greater in those with HBV markers. While sporadic HBV infection in such a setting is common, it is not a great health hazard.