Experimental studies of nephrotoxic nephritis in animals have demonstrated many of the underlying pathogenetic mechanisms involved in human glomerulonephritis. The first part of this thesis reviews the literature with regard to the factors and mechanisms responsible for the causation of glomerulonephritis and those factors which determine severity and allow for its persistence. In order to study these mechanisms, a reproducible model of nephrotoxic nephritis (NTN) in rabbits was established. This disease is produced by the intravenous injection of nephrotoxic serum, raised in sheep, containing antibodies to rabbit glomerular basement membrane. In a second group of experiments, the induction of an acute phase response in rabbits by the subcutaneous injection of a local irritant was studied. It was characterised by changes in C-reactive protein (CRP), the third component of complement (C3), fibrinogen and polymorphonuclear leucocytes (PMN' s), however renal function and renal histology were not affected in any way. When the so called heterologous phase of NTN in rabbits was induced during maximal acute phase stimulation, it did not result in enhancement of injury when compared to unstimulated rabbits. A similar situation was found during 7 the autologous phase, where a superimposed acute phase stimulus again did not cause enhancement of injury. Despite the fact that CRP has been shown to localize in injured tissue under certain circumstances, it did not fix in the kidneys of rabbits with induced NTN, nor in human kidneys affected by a variety of glomerulonephritides. The reasons for the variation in severity of disease during the autologous phase were analysed. Rabbits which developed severe glomerulonephritis during· the autologous phase of NTN characteristically produced high titres of rabbit anti-sheep antibody early in the disease when compared to those which did not develop injury, even though they might ultimately have developed the same titre of antibodies. One of the major determinants of injury appears to have been the rate at which antibody bound to the glomerular basement membrane (GBM). Studies during the heterologous phase confirmed this impression when it was shown that nephrotoxic globulin (NTG) given slowly, produced less injury than when the same amount was infused fast. This observation also suggested that the effector function of immunoglobulin decays much more rapidly than was previously thought. Finally, a mechanism of "protection" from injury by circulating antibody was shown by the phenomenon of "saturation" of available antibody binding sites on the 8 glomerular basement membrane which appeared to limit the deposition of further circulating antibody thus preventing further damage. These studies have advanced the understanding of human glomerulonephritis by illustrating how variation in immune responsiveness may contribute to the development of disease. They suggest that something inherent in infections, other than merely an acute phase response, is responsible for infection induced relapse in certain cases of allergic renal disease. By demonstrating a mechanism of “protection" from the effect of circulating antibodies, these studies suggest that therapy directed towards blocking of available antigenic sites on the human glomerular basement membrane, may become a therapeutic reality.