The clinical value of an enzyme-linked immunosorbent assay for the detection of immunoglobulin M (IgM) antibody to hepatitis B core antigen (anti-HBc IgM) was evaluated by testing serum samples from the following groups of patients: (a) 27 individuals who had been diagnosed as having acute hepatitis B virus (HBV) infection, (b) 29 hepatitis B surface antigen (HBsAg) carriers, (c) 6 subjects with acute non-B hepatitis, and (d) 10 HBsAg-negative but anti-HBc-positive subjects who were suspected of being index cases for the intimate transmission of HBV. Whereas 24 of the 27 individuals with presumed acute HBV infection exhibited anti-HBc IgM, only 2 of 29 HBsAg carriers were found to be positive. Hepatitis B surface antigen persisted during an 8-mo observation period in 3 anti-HBc IgM-negative subjects with acute HBsAg-positive hepatitis. Before anti-HBc IgM testing, it was considered that these cases had evolved to the HBsAg carrier state. However, the regular demonstration of anti-HBc IgM in acute type B hepatitis, as well as the failure to detect this antibody in the majority of HBsAg carriers, led to reclassification of these cases as probable instances of acute non-A, non-B or delta-agent hepatitis superimposed on the HBsAg carrier state. Through additional testing, the diagnosis of non-A, non-B (NANB) infection was confirmed in 2 of these cases, and delta-agent infection was identified in the third. None of the non-B hepatitis cases exhibited anti-HBc IgM. However, 5 of the 10 suspected type B index cases were anti-HBc IgM-positive, indicating that they were very recently infected and most likely had infected their cohabiting sexual partners. The results from this study indicate that testing for anti-HBc IgM may improve serodiagnostic accuracy when acute NANB and delta-agent hepatitis occur in previously unrecognized HBsAg carriers. Moreover, it may be a useful test in defining potential high risk sources of exposure to HBV.