The current work aimed at tracing the causative components for soybean-induced enteritis in Atlantic salmon (Salmo salar L.). Soybean molasses was subjected to phase separation using n-butanol. Three subtractions were obtained as follows: butanol phase, precipitate, and water phase. The biochemical composition of soybean molasses and the obtained subtractions were analyzed in detail: Protein, fat, and ash were quantified according to standard methods. Sucrose, raffinose, and stachyose were quantified using high-performance anion-exchange chromatography. Soyasaponins were quantified using reverse-phase high-performance liquid chromatography. Finally, sodium dodecyl sulfate-polyacrylamide gel electrophoresis was used to evaluate the size distribution of the proteins present in each fraction. Molasses and the different subfractions were thereafter fed to Atlantic salmon in two successive fish trials. The level of intestinal inflammation was evaluated by light microscopy using a semiquantitative scoring system. Histological assessments revealed that Atlantic salmon fed a combination of butanol phase and precipitate displayed significant enteritis. Atlantic salmon fed the water phase displayed normal intestinal morphology. The causative components for soybean-induced enteritis withstand butanol treatment and prolonged heating at 70°C. Sucrose, raffinose, stachyose, nor soybean proteins larger than 10 kDa induce enteritis alone. Soyasaponins, or components that follow the same solubility pattern, trigger the inflammatory reaction. We therefore suggest that soybean-induced enteritis in Atlantic salmon is induced by soyasaponins alone or by soyasaponins in combination with other factors, e.g., antigenic soybean proteins or the intestinal microflora.