Recent studies have demonstrated that the nephritogenicity of antibodies to dsDNA and nucleosomes confers to binding of glomerular membrane-associated nucleosomes, and not to cross-reacting glomerular antigens. There is no known parameter that determines antibody pathogenicity aside from specificity for dsDNA/nucleosomes, and systemic lupus erytheomatosus (SLE) patients may have high titer anti-dsDNA antibodies irrespective whether they have lupus nephritis or not. One parameter may be antibody affinity, as theoretically only high affinity antibodies may bind in vivo in a stable way. This was analyzed in (NZB × NZW)F1 mice with full-blown lupus nephritis. These mice had serum antibodies to dsDNA, and IgG autoantibodies bound in situ in glomerular membrane-associated electron dense structures as determined by immune electron microscopy (IEM). Intrinsic affinity of purified circulating and glomerular IgG anti-dsDNA antibodies was determined by surface plasmon resonance. The results demonstrate that affinity of glomerular-bound anti-dsDNA antibodies was higher than for those in circulation. However, affinity of glomerular in situ-bound antibodies from different mice varied considerably, from K_D in the range from 10^{− 8} to 10^{− 13}. These results indicate that antibody affinity is not a decisive pathogenic factor, but rather that availability of chromatin fragments may be the factor that determines whether an anti-dsDNA antibody binds in glomeruli or not.