Goodpasture's syndrome (GPS) is an autoimmune disease characterized by pulmonary hemorrhage, glomerulonephritis and anti-glomerular basement membrane (GBM) antibodies. The alpha(3) noncollagenous domain (NC1) of type IV collagen [alpha(3)(IV)] is the pathogen. The disease is T-cell-dependent; thus linear peptides initiate the autoimmune process. Studies in a rat model of GPS, experimental autoimmune glomerulonephritis (EAG), have shown that the carboxy-terminal 36 amino acids (purportedly the pathogenic epitope) are not responsible for disease induction. More recent studies implicate the amino terminus of alpha(3)(IV)NC1. Finding the nephritogenic epitope(s) is crucial in the understanding of the disease and for treatment. Because alpha(3)(IV)NC1 contains the antigens that induce GN in rats and human beings, we hypothesized that regions of the alpha(3)(IV)NC1 other than the carboxy terminus were responsible for disease. We investigated overlapping peptides spanning the entire NC1 domain of the alpha(3)(IV) chain N-terminal to the 36-mer (Goodpasture epitope) using the EAG rat model. Most peptides elicited antibody responses exclusively to themselves but not to native GBM. T-cells from GBM-immunized rats proliferated in vitro after stimulation with peptides 6, 8, 14, and 15, 24-mer and 23-mer. Fifteen percent of peptide 8 and peptide 14 rats had mild glomerulonephritis. In none of the animals immunized with other peptides did glomerulonephritis develop. These data suggest that conformation-dependent sites, posttranslational modification, multiple epitopes, concomitant antibody formation, or other disturbances are important in the ability of alpha(3)(IV)NC1 to induce EAG in rats and may also be important in the induction of GPS in human beings.