Abstract There was evidence that perinuclear antineutrophil cytoplasmic antibodies (pANCA) in autoimmune liver diseases react with human beta-tubulin-5 (TBB5). Here, we reevaluate the specificity and clinical relevance of anti-TBB5 antibodies. Patients with untreated autoimmune hepatitis (AIH; n = 53), AIH under immunosuppressive therapy (AIH-IS; n = 125), primary sclerosing cholangitis (PSC; n = 40), primary biliary cholangitis (PBC; n = 250), nonautoimmune liver diseases (n = 158), inflammatory bowel diseases (IBD; n = 30), and healthy individuals (n = 62) were tested by enzyme-linked immunosorbent assay for IgG- and IgA-antibodies against recombinant human TBB5. pANCA were detected by immunofluorescence test. Sera were absorbed with TBB5 coupled to cyanogen bromide-activated sepharose. Prevalence and reactivity of IgG anti-TBB5 were significantly higher in patients with untreated AIH (68%; arbitrary units [AU] median: 369) than in PSC (28%; AU median: 84, P < 0.001), other liver diseases (14%; AU median: 185, P < 0.0001), IBD (3%; AU median: 111, P < 0.0001), and healthy controls (3%; AU median: 135; P < 0.0001). Anti-TBB5 did not correlate with pANCA, and immunoprecipitation with TBB5 did not abolish pANCA reactivity. In untreated AIH, anti-TBB5-reactivity was significantly higher than in AIH-IS. Transaminases decreased under IS preferentially in anti-TBB5-negative patients. There was no correlation between anti-TBB5-reactivity and histological stages. IgA-anti-TBB5 was mainly found in alcohol-associated liver disease (ALD; 39%). Our data do not support TBB5 as an autoantigenic target of pANCA. However, IgG-anti-TBB5 showed high specificity for (untreated) AIH. While they did not correlate with histological and laboratory parameters, their presence may indicate a poor response to IS.