Objective: a part of patients with primary biliary cirrhosis (PBC) has anti-human carbonic anhydrase II (CA II) autoantibodies, although several contradictional reports followed. Since immunization of mice with CA II results in cholangitis in a susceptible strain, PBC with anti-CA II antibody may have distinct clinical features. Thus, we tested the sera of patients with PBC for anti-CA II antibodies and compared clinical characteristics of PBC patients with and without anti-CA II antibodies in Japanese patients. Methods: anti-CA II antibodies were detected in nine of 50 (18%) PBC patients by immunoblotting. The evaluation of these patients included various clinical parameters, autoantibodies, and immunological backgrounds. Results: the levels of serum liver tests and the prevalence of serum anti-mitochondrial antibody (77.8 vs. 92.7%) were not different between the patients with and without anti-CA II antibody. However, the prevalence of anti-nuclear antibody (ANA) was significantly higher in the patients with anti-CA II antibody than that in the patients without anti-CA II antibody (66.7 vs. 25.6%, P=0.044), although their mean titers were not statistically different. Association of Sjøgren's syndrome tended to be more frequent in the patients with anti-CA II antibody than those without it (33.3 vs. 14.6%, P=0.327). Studies of HLA class I allotype revealed that three of five (60.0%) patients with anti-CA II antibodies and one patients from 34 (3.0%) patients without anti-CA II antibodies had HLA B51 allotype; the difference in the prevalence of this allotype was significant (P=0.004, Pc=0.01), and the prevalence of other HLA class I and HLA DR allotypes was similar between the patients with and those without anti-CA II antibody. Administration of ursodeoxycholic acid (600 mg per day) was accompanied by change in liver tests in a similar way between the two patient groups. Conclusions: These results suggest that, although clinical features are not distinctive, PBC patients with anti-CA II antibody may have a genetic background, which may contribute to a susceptibility to immune-mediated cholangitis.