1988. Genetic interactions among four genes for resistance to stem rust in bread wheat. Genome, 30: 182 -185. Near-isogenic lines of bread wheat (Triticum aestivum L.) carrying the genes Sr7a, Sr8a, Sr9b, and S r l l for stem rust (Puccinia graminis f. sp. tritici Eriks. and Henn.) resistance had been developed previously using nine backcrosses to the cultivar Marquis. The lines were intercrossed to produce the six possible two-gene combinations and three of the four possible three-gene combinations. The objective was to test for additive, modifying, or residual effects of the genes. The nine combinations, the four near-isogenic parents, 'Marquis', and three near-isogenic lines with only five backcrosses to 'Marquis' were tested. Seedling tests with five races were done in a growth room and field tests were done with two races. Relatively few interactions among the genes were detected in the seedling tests. All involved small additive effects when Sr7a was combined with either Sr8a or Sr9b. Interactions were much more common in the field tests. With race 56, the three two-gene combinations of Sr8a, Sr9b, and S r l l showed significant additive effects. With race 15B-1, there were four cases of an ineffective gene modifying the effect of an effective gene. In addition, the data indicated that Sr9b and possibly S r l l had residual effects against race 15B-1. The reaction of a genotype to stem rust in the field clearly involves a complex interaction of effective and ineffective genes. among four genes for resistance to stem rust in bread wheat.Genome, 30 p: 182 -185. Des lignkes de blk panifiable (Triticum aestivum L.), pratiquement isogkniques, porteuses des gknes Sr7a, Sr8a, Sr9b et S r l l pour la rksistance B la rouille de la tige (Puccinia graminis f. sp. tritici Eriks. et Henn.) ont kt6 produites antkrieurement au moyen de neuf rktrocroisements avec le cultivar Marquis. Ces lignkes furent croiskes entre elles en vue de produire les six combinaisons possibles de deux gknes et trois des quatres combinaisons possibles de trois gknes. L'objectif ktait de vkrifier les effets additifs, modificateurs ou rksiduels des gknes. Les tests de rksistance B la rouille des tiges ont kt6 effectuks avec les neuf combinaisons, les quatre parents presque isogkniques, le cv. Marquis et trois lignkes presque isogkniques avec seulement cinq rktrocroisements avec le cv. Marquis. Les tests furent faits en chambre de croissance sur des plantules, avec cinq races de P. graminis f. sp. tritici et, en plein champs, avec seulement deux races. Les tests de plantules n'ont rkvklk que relativement peu d'interactions entre les gknes. Toutes les interactions impliquaient de petits effets additifs lorsque le gkne Sr7a ktait combin6 soit avec le Sr8a ou le Sr9b. Les tests de plein champs ont rkvklk des interactions beaucoup plus marqukes. Ainsi, avec la race 56, les trois combinaisons de deux gknes, Sr8a, Sr9b et S r l l ont montrk des effets additifs significatifs. Avec la race 15B-1, il y a eu quatre cas oti un gkne inefficace a modifik l'effe...