This study was conducted at the Experimental Farm of the Faculty of Agriculture, Ain Shams University, at Shalakan, Kalubia Governorate, Egypt during the three successive growing seasons 2012/2013, 2013/ 2014 and 2014/2015 to study the relative response to selection in three bread wheat crosses for two cycles started in F3 using pedigree selection procedure for improving grain yield and some other agronomic traits, i. e. days to heading, plant height, spike length, number of spikes/ plant, number of spikelets/ spike, number of kernels/ spike and 1000-kernel weight. The results obtained revealed the presence of sufficient genetic variability among selected families or genotypes regarding the studied traits and a great response to selection can be expected from the progenies of these crosses. Genotypic coefficient of variation values were near those of phenotypic coefficient of variation for all traits, indicating minor role for the environment on variation among genotypes. This may interprets the high values of heritability and expected genetic gain of selection. The highest estimates of phenotypic and genotypic coefficient of variation were shown by number of spikes/ plant and grain yield/ plant while, the low estimates were shown by days to heading and no. of spikelets/ spik across the F3, F4 and F5 generations. The heritability estimates in broad sense were; low for number of spikes/ plant, moderate for spike length, number of spikelets/ spike and 1000-kernel weight and high for days to heading, plant height, number of kernels/ spike and grain yield/ plant across the three generations. The promising lines; 7, 8, 10, 19, 24, 27, 30 and 31 exhibited the best performance in F5 –generation and surpassed most check cvs. in grain yield and most yield components. Thus, these promising lines could be used in breeding programs for improving wheat grain yield. Grin yield was positively and significantly correlated with each of spike length, no. of spikelets/ spike, no. of spikes/ plant, no. of kernels/ spike and 1000-kernel weight at phenotypic and genotypic levels and with each of days to heading and plant height at genotypic level only. Thus, these traits could be considered as selection criteria in wheat breeding programs for yield improvement.