To elucidate the questions of why not all patients with hepatitis B virus (HBV) infection develop HBV membranous nephropathy (HBVMN), we first measured serum HBe circulating immune complex (CIC) during the acute nephrotic phase of HBVMN and in HBV carriers. We found that the level of HBe CIC was low in the HBVMN patients and absent either in HBsAg+/HBeAg+ patients without HBVMN or HBsAg+/HBeAg- asymptomatic carriers. Second, we needed to characterize the cellular immune response to HBV in patients with HBVMN. However, lack of a suitable autologous effector/target cell system makes a precise study of HBVMN pathogenesis difficult. In the present study, we established a model system by using autologous HBcAg-expressing Epstein-Barr-virus-immortalized lymphoblastoid cell lines (LCL) as stimulator/target cells. Both proliferative response after stimulation with HBcAg and cytotoxic activity against autologous HBcAg-expressing LCL of the peripheral blood T cells obtained from the HBVMN patients and HBsAg carriers could be measured. Using autologous HBcAg-expressing LCL as stimulator/target cells for the study of HBcAg-specific cytotoxic T lymphocytes, we found that HBVMN patients had lower cytotoxic activity than did both HBV carriers and HBsAg-/HBsAb+, HBeAg-/HBeAb+ children. From the in vitro cytokine production study of peripheral blood T cells after stimulation with HBcAg, we found that T-helper-cell-1-related IL-2 and IFN-gamma productions were very low in HBVMN patients but T-helper-cell-2-related IL-10 production was higher in HBsAg+/HBeAg+ patients with HBVMN than in those without HBVMN. Based on these findings, we conclude that HBVMN children seem to have an inadequate cellular immune response to HBcAg.