Laboratory-prepared and commercially obtained fluorescein-labeled rabbit antihuman IgG were compared in performing the antimitochondria antibody (AMA) assay. Identical results were obtained using either of the fluorescent antisera at protein concentrations of 1.5 mg/ml and 1:10 dilutions of patients' sera. Positive AMA tests with either antisera were observed in each of 7 patients with primary biliary cirrhosis (PBC), 2 of 83 patients with miscellaneous hepatic diseases, and in 1 of 24 patients with extrahepatic biliary obstruction (EBO) of 2-24 weeks duration. However, when undiluted commercial fluorescent antiserum (15.8 mg protein/ml) was substituted in the assay, sera from 11 of 23 AMA-negative patients with EBO and 12 of 15 with miscellaneous liver diseases demonstrated an atypical fluorescence located primarily along the periphery of the rat renal tubules. Thus, if the conjugated antibody is not adjusted to an optimal protein concentration, this atypical fluorescence could be interpreted as a positive AMA test and lead to diagnostic error.