The indirect fluorescent-antibody (IFA) method for diagnosis of toxoplasmosis is widely used and is considered to be as specific as the Sabin-Feldman dye test. After observing a patient with systemic lupus erythematosus (SLE) who had a positive toxoplasma IFA test but a negative dye test, we studied sera with high titers of antinuclear antibodies from 16 SLE patients and from 2 with rheumatoid arthritis for Toxoplasma antibodies in the immunoglobulin G and M (IgG and IgM) IFA tests and the dye test. Results of these tests were compared with titers of antinuclear antibodies, precipitating antibodies to single-strand deoxyribonucleic acid (DNA), and binding antibodies by use of DNA labeled with (3)H-actinomycin D. Of 18 patients, 11 had IgG and 4 had IgM IFA Toxoplasma antibodies; only 2 had antibodies detectable in the dye test. The immunofluorescence patterns in the Toxoplasma IFA test were indistinguishable from those obtained in patients with toxoplasmosis without antinuclear antibodies. Absorption of SLE sera with DNA did not result in a decrease in Toxoplasma IFA titers. When SLE sera were absorbed with live T. gondii, a marked drop in IgG IFA titer was observed as well as a decrease in titers of antinuclear antibodies and (3)H-DNA binding. Treatment of Toxoplasma cells with deoxyribonuclease and ribonuclease did not decrease their fluorescence. These results suggest that T. gondii nuclear antigens can absorb antinuclear antibodies but do not have exposed substrates for deoxyribonuclease. Tests in which organisms containing "nuclear" antigens for IFA detection of antibodies to these organisms are used may result in "false-positives" with sera containing antinuclear antibodies.