Serum anti-mitochondrial antibodies (AMA) are the serological hallmark of primary biliary cirrhosis (PBC), yet up to 15% of PBC sera are AMA negative at routine indirect immunofluorescence (IIF) while being referred to as "probable" cases. The diagnostic role of PBC-specific antinuclear antibodies (ANA) remains to be determined. We will report herein data on the accuracy of new laboratory tools for AMA and PBC-specific ANA in a large series of PBC sera that were AMA-negative at IIF. We will also provide a discussion of the history and current status of AMA detection methods. We included IIF AMA-negative PBC sera (n=100) and sera from patients with other chronic liver diseases (n=104) that had been independently tested for IIF AMA and ANA; sera were blindly tested with an ELISA PBC screening test including two ANA (gp210, sp100) and a triple (pMIT3) AMA recombinant antigens. Among IIF AMA-negative sera, 43/100 (43%) manifested reactivity using the PBC screening test. The same test was positive for 6/104 (5.8%) control sera. IIF AMA-negative/PBC screen-positive sera reacted against pMIT3 (11/43), gp210 (8/43), Sp100 (17/43), both pMIT3 and gp210 (1/43), or both pMIT3 and Sp100 (6/43). Concordance rates between the ANA pattern on HEp-2 cells and specific Sp100 and gp210 ELISA results in AMA-negative subjects were 92% for nuclear dots and Sp100 and 99% for nuclear rim and gp210. Our data confirm the hypothesis that a substantial part of IIF AMA-negative (formerly coined "probable") PBC cases manifest disease-specific autoantibodies when tested using newly available tools and thus overcome the previously suggested diagnostic classification. As suggested by the recent literature, we are convinced that the proportion of AMA-negative PBC cases will be significantly minimized by the use of new laboratory methods and recombinant antigens.