This investigation was made to evaluate three bread wheat hybrids and their six populations generated from each of them under normal and late sowing dates. The three hybrids were obtained where hybrid 1 was a result of crossing (Late × Late) cultivars; hybrid 2 was a result of crossing (Early × Early) cultivars and hybrid 3 was a result of crossing (Late × Early) ones. The early parental in breed lines included: Line1, Line2 and Line3 while, thelate parental onesincluded: Misr2 and Gemmeiza9. The two sowing dates were: normal sowing on 29 November 2011 and the late sowing on 29 December 2011. Therefore, the total genetic materials were obtained from the three hybrids included 18 genotypes which were evaluated at Sakha Agricultural Research Station Farm, Wheat Section, Egypt. Earliness traits showed that late sowing date required fewer days than normal sowing date for days to heading; days to maturity and grain filling period.  This trend was present for all populations of the three hybrids. It appeared that late sowing shorten the life span of plants. Yield component characters which included: plant height; 100- kernel weight and grain yield/ plant showed the advantage of hybrid 1 for plant height and grain yield/plant than the other two hybrids. These two characters showed a modest amounts of heterosis for the three hybrids. Nature of gene action which estimated from the six populations and evaluated using the scaling test analysis  proposed by Mather and Jinks (1971). The results indicated the importance of additive genetic variance than dominance genetic variance one for most studied traits. These results were also supported by the results obtained from the scaling test analysis as the quantities A,B and C indicated the adequacy of the additive dominance model.