Gene pyramiding is a breeding strategy whereby host resistance genes are combined together with the objective of prolonging their usefulness in crops such as wheat (Triticum aestivum) for resistance to leaf rust caused by Puccinia recondita f.sp. tritici. When genes are combined they often give reactions different from those given by each component gene alone. Effects of gene combinations in lines Lr13 + Lr34 (T34-13), Lr13 + Lr37 (T13-37) and Lr34 + Lr37 (T34-37) were compared with those of the single gene lines CT263 (Lr13)(T13), RL6058 (Lr34)(T34), RL6081 (Lr37)(T37) and the leaf rust susceptible control, Thatcher. Infection types on plants infected with pathotypes UVPrt2 or UVPrt13 in the glasshouse, and disease severity in the field, demonstrated higher levels of resistance in the combination lines T13-37 and T34-37 than in the lines with the individual genes. The absence of sporulating uredinia in these combination lines prevented quantitative measurements of components such as latent period. In the T34-13 line, no increased resistance to pathotype UVPrt13 was apparent from assessment of the infection types in the glasshouse. Precise measurements of its resistance components showed, however, that it had a longer latent period and smaller uredinia and its resistance was highly effective in the field. There was variation in leaf rust severity amongst sister lines containing both Lr13 and Lr34, suggesting that increased resistance in T13-34 may not be controlled solely by these two genes themselves. Development of fungal structures, and the incidence and area displaying a hypersensitive reaction, were assessed using UV-1A and B-2A fluorescence microscopy filter combinations. Significant restriction of fungal growth during early postinfection stages occurred in the gene combination lines T34-13, T13-37 and T34-37. Colony size in these lines was also significantly reduced compared with that in the single gene lines T13, T34, T37 and the leaf rust-susceptible Thatcher, when either or both pathotypes possessed avirulence for one of the Lr genes. In the compatible T13-34/ UVPrt13 interaction no clear histological evidence of resistance enhancement was observed. The hypersensitive reaction in Lr37 alone or in combination and with either pathotypes, and Lr13 alone or in combination and with UVPrt2 indicated that the major component of the resistance mechanism is posthaustorial. Lr34 had the lowest hypersensitive index and even less hypersensitivity was observed in the Lr34/Lr37 combination line than in Lr37 alone. There was more prehaustorial abortion of infection structures in the line T34-37 than in T34 and T37 with isolate UVPrt2 but not with UVPrt13. It did not appear that abortion of infection structures was a major component of the resistance studied.