In this study, twenty fife bread wheat genotypes were tested under eight environments (2 growing seasons × 4 water stresses) on physiological and agronomic traits by using Eberhart and Russell (1966), Perkins and Jinks (1968) and Freeman and Perkins (1971) models to study stability and adaptability of the genotypes. Analysis of variance showed highly significant variations among the tested bread wheat genotypes for all studied traits. The results indicated that Line 9 (13.21 ardab/feddan) beside Sids 1 (13.19 ardab/feddan) and Misr 2 (11.91 ardab/feddan) cultivars were observed as most stable and widely adapted across environments, surprised mean performance for grain yield than grand mean, regression coefficient (biEberhart and biFreeman) equal or near to one, BiPerkins equal or near to zero and mean square deviation from regression (S2diEberhart and S2diFreeman) nonsignificant, according to previous models. Moreover, Line 9 showed also stability for number of spikes/plant, 1000 kernel weight and chlorophyll concentration under various environments, it was considered to be superior for grain yield across different environments. Meanwhile, the highest yielding genotype (Line 20) showed stability only under favorable conditions (surpassed the grand mean by 15.58 ardab/feddan, biE and biF > 1, Bi > 0 and S2diE and S2diF non-significant), it could be promoted to breeding program with a stable genotype (Line 9) over different environments to obtain a stable variety across environments with a high yield.