Drought stress is one of the most important environmental stresses that affect yield of bread wheat strongly. Considering that the yield achieved with the interaction of a large number of plant traits and environment, so the morphological and agronomical traits, can serve as selection criteria under environmental conditions. By genetic analysis and estimation of inheritance and mode of gene action of yield-related traits, we can determine effective breeding methods specifically under drought stress conditions. For this reason, 2 crosses Kavir×Shiraz and Azar2×Shiraz were performed (Shiraz is a susceptible cultivar, while Azar2 and Kavir are tolerant cultivars). F1, F2, BC1, BC2 generations obtained and along with their parents were planted in a randomized complete block design with three replications under drought stress condition. Various agronomic traits such as number of tiller, length of plant, length of main spike, length of peduncle, weight of plant, weight of main spike, weight of tiller, number of grains in main spike, 1000-grain weight, date of maturity and harvest index were measured. Results of weighted analysis of variance showed significant difference between generations for all traits. Generation mean analysis was performed. In more traits, in addition to additive and dominant effects, epistasis effects and additive× additive effects were added to the model. In traits of weight of plant, weight of stem, number of grains in main spike, and date of maturity, dominant effects was more than additive effects. Therefore, breeding of these traits is not useful and hybridization is better. For 1000-grain weight that additive effects was more than dominant effects, can use various selective methods for breeding. In traits of 1000-grain weight, number of grains in main spike, date of maturity, both of selective-based and hybrid-based methods are useful. Mean of heritabilities in stress and non-stress condition were 0.592 and 0,517. Estimation of broad, narrow sense heritabilities and mode of gene action indicated that selection for improvement of traits studied in stress condition will be suitable and traits such as length of plant and spike, weight of shoot, date of maturity, and harvest index can be used as indirect selection criteria for improving grain yield under drought stress condition. Genetic control of number of tiller, length of plant, length of peduncle and number of grains in main spike was not similar.