Associations of different assays for antibodies to C1q (anti-C1q) and to dsDNA (anti-dsDNA) and of complements C3 and C4 with disease activity in patients with systemic lupus erythematosus (SLE) were studied. The clinical manifestations of 223 SLE patients were recorded, and the disease activity was assessed by the SLEDAI score. Anti-C1q were determined by two enzyme-linked immunosorbent assays (ELISA) and anti-dsDNA by a radioimmunoassay (RIA), a Crithidia immunofluorescence (IF) assay and three ELISA assays using human telomere DNA, plasmid DNA circles, or calf thymus DNA as antigens, respectively. Complement C3 and C4 were determined by nephelometry. Control sera were obtained from 98 blood donors. In patients with SLE, the prevalence of anti-C1q was 17-18% and that of anti-dsDNA was 36-69%. Anti-C1q, anti-dsDNA, and complement C3 and C4 correlated well with the overall activity of SLE (r = 0.323-0.351, 0.353-0.566, and -0.372-0.444, respectively; P < 0.001). Sensitivity, specificity, positive predictive value, and negative predictive value for active lupus nephritis among SLE patients were 40-44, 92, 29, and 91-92% for anti-C1q and 48-68, 29-66, 11-16, and 86-91% for anti-dsDNA, respectively. Patients with active nephritis had higher levels of anti-C1q and lower levels of C3 and C4 than patients with inactive nephritis (P = 0.003-0.018). The corresponding associations of anti-dsDNA were somewhat weaker (P = 0.023-0.198). Hematological parameters reflecting disease activity correlated clearly better with anti-dsDNA and complement C3 and C4 than with anti-C1q. Anti-C1q is inferior to anti-dsDNA as a diagnostic test in SLE and in the evaluation of overall clinical activity of the disease. Anti-C1q together with complement C3 and C4 may offer useful additional information to monitor lupus nephritis activity. There are no practical differences between different assays for anti-C1q and anti-dsDNA.