IgM-anti-HBc and IgG-anti-HBc serum titers were determined by indirect immunofluorescence in a prospective longitudinal study of 50 patients with hepatitis B, 43 of whom recovered completely. 37 of the recovered patients and all 7 non-recovering patients were followed up for a median of 5 years. Five of the non-recovering patients were followed up from the initial acute stage of the disease. IgM-anti-HBc was present in the acute stage in 39/43 of the recovery patients. The median maximal titer, 1:1000, was reached during the week before peak SGPT. It always disappeared in recovering patients within a median period of 5 weeks after peak SGPT. IgG-anti-HBc was present in all 43 recovering patients in the acute stage of disease with a median maximal titer of 1:1000, maintained for at least 10 weeks. After 5 years, 28 of 37 recovered patients were still IgG-anti-HBc positive with a median titer of 1:200. All non-recovering patients showed persistent IgM as well as IgG-anti-HBc positivity. In the acute stage the medians of the maximal titers were 1:100 for IgM-anti-HBc and 1:1000 for Igg-anti-HBc. After 5 years they were 1:100 for IgM and 1:10000 for IgG-anti-HBc. The presence of IgM-anti-HBc in a preceding study was considered to be a marker of hepatitis B virus replication. From this study no evidence can be obtained to support the view that the titer level of anti-HBc is reliable in the differentiation between infectious anti-HBc positive blood, as there was no difference (p = 0.4) between the number of patients with an anti-HBc level of 1:1000 after at least five years, who had recovered (9/28) and who had not recovered (3/7).