The histological findings in patients with primary biiary cirrhosis have been well‐defined and are often used in the clinical staging of disease. However, it has only been with the development of reagents that phenotypically characterize the lymphoid infiltrate that attempts have been made to correlate pathophysiology with immune effector populations. Indeed, the inflammatory hepatic lesions in primary biliary cirrhosis have been described as containing CD4‐positive and CD8‐positive T cells. Less clear, however, have been the T cell receptors in these lesions. Further, the data on immunoglobulin deposits in hepatic lesions have been less well‐defined; this deficit may be a result of the quality of polyspecific sera and difficulties in background. To address these issues, we have used a battery of well‐defined monospecific and polyspecific reagents to phenotypically define the occurrence of lymphoid cells in the livers of patients undergoing transplantation. Furthermore, we have defined these same markers on T cell lines derived from liver, regional lymph node and peripheral blood. The predominant cell type in the mononuclear infiltrate is the CD3+, CD4+ T lymphocyte bearing the T cell receptor αβ. T cell lines from the same patients demonstrate similar findings. Of special importance, however, was the detection of CD20+ B cells and Ig+ cells in the lymphoid infiltrate. Indeed, we also readily demonstrated the presence of immunoglobulin on the surface of biliary epithelium. These data suggest that mechanisms involved in the pathophysiology of primary biliary cirrhosis may include both T cell and antibody mechanisms. The results also underscore the need to develop a functional, and not just a phenotypical, assay of the inflammatory infiltrate. (HEPATOLOGY 1990;12:306–313).