This study aimed to investigate the response of lentil genotypes to different water regimes, providing guide lines, through partitioning the variation, for a selection program for adaptation to irrigated conditions. The research was divided into two main areas; 1) The overall variation in the crop was partitioned into genotypic, environmental and genotype-environmental components in an analysis of adaptation over seasons, irrigation regimes and locations; 2) The genotypic variation was partitioned into its various genetic components in an inheritance study using the dial lei mating system. Pronounced progress should be expected from selection for number of pods/plant, 100 seed weight and straw yield/plant traits, which showed high estimates of h(^2)(_b.s), C.G.V. and G.S. The two former traits correlated strongly and positively with seed yield, which allowed their use in indirect selection for seed yield. The 35 genotypes used in this study showed wide genetic diversity, allowing selection of high yielding genotypes under irrigation. Environmental variation in water supply, temperature and soil type was found to exert a profound effect on variation in characters measured. This suggests the possibility of raising yield levels through improved management practices. In this study, irrigation repeated twice increased seed yield by 19% over no irrigation, at the same location, and increased the yield by 300% in comparison with a dry location. Seed protein quality was influenced by environments and genotypes. Electrophoretic studies showed that the number and position of the bands could be used to identify genotypes. Four genotypes showed response to irrigation and could be recommended as promising entries. An anatomical study showed that large air spaces formed in the roots of a responsive genotype:, which could be used as a selection criterion for positive response to irrigation. Seed yield/plant exhibited 31.8% heterosis and showed a predominant role of non-additive genetic variance. Due to the significance of the non-additive effect, the superior F(_1)'s may be expected to throw out desirable transgressive segregants, provided that the complementary genes and epistatic effects included in the non-additive component are coupled in the same direction to maximize seed yield. Five F(_2) crosses showed superiority in seed yield and SCA effects. These crosses should be carried forward in lentil breeding programs.