OBJECTIVE: To evaluate antinuclear antibody (ANA) tests in established cases of systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA) by indirect immunofluorescence microscopy (F-ANA) and enzyme-immunoassays detecting antinucleosomal antibodies (ANSA-EIA). METHODS: Sera from 50 patients with SLE and 65 patients with RA were analyzed regarding abnormal concentrations of F-ANA (serum dilution >/= 1:200 = 95th percentile among 300 healthy blood donors). The sera were also analyzed with 2 commercial ANSA-EIA kits. RESULTS: An abnormal F-ANA titer occurred in 76% of the SLE sera compared to 23% in RA, and was not related to present use of antirheumatic drugs. At dilution 1:50, 84% of the SLE sera were F-ANA-positive compared to 20% of healthy women. Forty percent and 56%, respectively, of the SLE sera tested positive in the 2 ANSA-EIA kits. By the most sensitive assay, 96% of the ANSA-positive SLE sera produced a homogenous (chromosomal) F-ANA staining pattern compared to 18% of the ANSA-negative SLE sera. Ten of the 15 F-ANA-positive RA sera (63%) generated homogenous F-ANA staining and 13 (20%) tested positive in the most sensitive ANSA-EIA, but with no correlation to the F-ANA staining pattern. CONCLUSION: The sensitivity of F-ANA at an abnormal titer was surprisingly low (76%) in established cases of SLE. ANSA occurred in 56% of the SLE sera, but also in a fair number (20%) of RA sera. Practically all ANSA-positive SLE sera were identified by chromosomal F-ANA staining. We conclude that the antigen-specific antinucleosomal EIA does not have high enough diagnostic specificity to justify use of this analysis for routine diagnostic purposes.