It is accepted that the main determinant of glomerular injury in experimental nephrotoxic nephritis is the administered dose of anti-glomerular basement membrane (GBM) antibody. However, there are other factors that can enhance the severity of such injury including small doses of bacterial lipopolysaccharide (LPS). In the present study, we have assessed whether preparations of anti-GBM antibody contaminated with different concentrations of endotoxin could influence the severity of glomerular injury in the heterologous phase of nephrotoxic nephritis. We have also examined the efficacy of different laboratory methods to isolate an endotoxin-free anti-GBM antibody, and to purify anti-GBM antibody preparations from endotoxin. Preparations of anti-GBM antibody (nephrotoxic globulin) isolated from nephrotoxic serum by the sodium sulphate precipitation method contained variable concentrations of endotoxin. Administration of these preparations in equal doses into clean rats, which had no established acute phase response, markedly aggravated the severity of glomerular injury. However, preparations contained less than 50 pg/ml of endotoxin appeared to have no significant effect on such injury. Furthermore, isolation of anti-GBM antibody from nephrotoxic serum by affinity chromatography, using Staphylococcus protein-A column, proved to be a reliable method not only for the isolation of an IgG (nephrotoxic antibody) free from other serum contaminants, but also for purification of endotoxin contaminated preparations of anti-GBM antibody. These observations have practical implications in studying models of nephritis as our results show that the glomerular injury, which is usually considered to be a sole function of the mass of antibody bound to GBM, is profoundly influenced by minor endotoxin contamination of the anti-GBM antibody.