Autoantibodies to RNA polymerases (RNAP) I, II, and III are reported to be highly specific for the diagnosis of scleroderma (systemic sclerosis, SSc). In the present study, the specificity of autoantibodies to RNAP I and III for SSc was confirmed by immunoprecipitation of 35S-labeled proteins. However, we report here the previously unrecognized production of anti-RNAP II autoantibodies by 9-14% of patients with SLE and mixed connective tissue disease/overlap syndrome. 12 out of 32 anti-RNAP II positive sera (group 1) immunoprecipitated a diffuse 220-240-kD band identified as the largest subunit of RNAP II whereas the remaining 20 (group 2) immunoprecipitated preferentially the 240-kD phosphorylated (IIo) form of the large subunit. After pulse labeling, group 1 sera immunoprecipitated only the 220-kD (IIa) RNAP II subunit, whereas the diffuse IIa/IIo band plus the 145-kD second largest RNAP II subunit (IIc) were immunoprecipitated after several hours of cold chase, suggesting that these sera recognized primarily the largest subunit of RNAP II. Group 2 sera recognized the IIc subunit after pulse labeling, and immunoprecipitated the IIc and IIo, but not the IIa, subunits after cold chase. Although it has been suggested that autoantibodies to RNAP II are usually accompanied by anti-RNAP I/III in SSc, all but one of the anti-RNAP II positive sera from SLE or mixed connective tissue disease/overlap syndrome patients, as well as most of the SSc sera, were negative for anti-RNAP I/III. Moreover, in contrast to previous reports suggesting that anti-RNAP antibodies rarely coexist with other SSc subset marker antibodies, anti-RNAP II antibodies were often accompanied by anti-Ku, anti-nRNP, or anti-topoisomerase I autoantibodies in the present study. We conclude that autoantibodies to RNAP II are not a specific marker for SSc, whereas autoantibodies to RNAP I/III are associated primarily with SSc. In addition, we have identified two distinctive patterns of RNAP II antigen recognition by autoantibodies, one of them characterized by specific recognition of the transcriptionally active (phosphorylated) form of RNAP II. The clinical significance of these different patterns remains to be determined.