Generation mean analysis for drought tolerance was studied in wheat cross-1 (Pavon-76 X Gemmeiza-7) and cross-2 (ICR-DH18 X Pavon-76). The genotypes were evaluated under control and drought stress (15% polyethylene glycol 6000) at germination and seedling stage for seven traits. The additive–dominance model is adequate for explaining the inheritance of root and shoot lengths under both treatments and root fresh weight under drought stress in cross-1, and shoot length in both treatments and root length under drought stress in cross-2, while being inadequate in the other traits in the two crosses treatments. The additive-dominance and epistatic interaction effects recorded for germination percentage, root and shoot fresh and dry weights in both crosses suggested postponement of plant selections till the later generations for plant traits with such type of gene action. Epistasis absence and the contribution of considerable additive genetic variance in root and shoot lengths and root fresh weight indicating that recurrent selection in early segregating generations could be effective to select wheat lines with enhanced early tolerance to drought stress. Only two SSR primers and two TRAP primer pairs generated polymorphic bands from the tested genotypes. Four positive molecular markers were detected for drought tolerance. The UPGMA clustering analysis revealed correlation between drought tolerance genotypes and the studied molecular markers. The markers identified herein would allow implementing marker-assisted selection to screen wheat segregating populations for drought tolerance.