Sera from 35 patients with biopsy-proven diffuse proliferative (WHO class IV) or membranous (WHO class V) lupus nephritis were analyzed for the presence and size of circulating immune complexes. Elevations of the C1q solid-phase assay (C1qSP) for immune complexes were found in sera from all patients with diffuse proliferative nephritis, with a mean +/- 1 SEM of 166.8 +/- 42.0 micrograms/AHG-equivalents/ml serum, and in 71.4% of the patients with membranous nephritis (83.1 +/- 26.7, p = 0.06). Using the WHO criteria for subclasses of membranous lupus nephritis, we also designated renal biopsies as nonproliferative (WHO classes Va and Vb) or proliferative (WHO classes IV and Vc). Employing the latter groupings, we observed significant differences between C1qSP results of patients with nonproliferative (30.3 +/- 8.8) and proliferative (172.8 +/- 36.8, p less than 0.001) lupus nephritis. These data suggest that the presence of C1q-binding material in serum is pathophysiologically related to proliferative glomerular lesions, and that levels of C1qSP binding reflect renal lesions in SLE patients. Sucrose density gradient ultracentrifugation was performed on each serum, and gradient fractions analyzed for C1qSP-binding and total IgG, using techniques to minimize losses of immune complexes. The predominant peak of C1qSP activity sedimented with the 6.6S monomeric IgG. The 6.6S C1q-binding IgG was increased only in 1 of 10 patients with membranous lupus nephritis without proliferative changes, and was elevated in 16 of 25 patients with proliferative lesions (WHO classes IV and Vc). A significant negative correlation was found between the presence of this C1q-binding material and subepithelial electron-dense deposits, suggesting that the presence of this material contributed to the absence of subepithelial immune deposits. Large-molecular-weight C1qSP-binding material was also present, mainly in sera from patients with proliferative lesions. Furthermore, highly positive correlations were found between immune deposits in interstitial blood vessels and peritubular areas, and the concentrations of C1qSP-binding IgG and rapidly sedimenting IgG in density gradient analysis. Overall, these findings are consistent with the hypotheses that circulating immune complexes contribute to the pathogenesis of glomerulonephritis and interstitial nephritis in patients with SLE, and that 6.6S C1q-binding IgG plays a role in the proliferative lesions of lupus glomerulonephritis.