Ten diallel crosses among five flax cultivars showing clear differences were made and evaluated at the Agricultural Experiments Desert Station, Faculty of Agriculture, Cairo University, in Wadi El-Natroon, El-Beheira Governorate, during three successive growing seasons starting with 2011/2014. The objectives were to estimate the mode of gene action in the inheritance and combining ability effects for studied traits. Combining ability variances were highly significant for all studied traits except plant height at F2 generation indicating the importance of additive and non-additive genetic variances in determining the performance of these traits. Since the ratio of GCA/SCA exceeded the unity, suggesting that additive was much larger and more important than non-additive gene effects in the inheritance of all studied traits. Data indicated great and significant genotypic differences in all studied traits. The averages of F1 crosses were generally higher than F2 in straw yield plant-1 and plant height, while the averages of F1 and F2 were identical in seed yield plant-1, No. of capsules plant-1, oil % and oil yield plant-1. However, crosses averages were higher than parents averages in F1 for straw yield plant-1, plant height, seed yield plant-1, and oil yield plant-1 by 11.83, 4.12 1.39 and 0.71%, respectively. This indicated the superiority of heterozygotes over homozygotes in flax performance. The best general combiners was P1 (Sakha 1) for all traits in F1 while good for straw yield plant-1, seed yield plant-1, No. of capsules plant-1 and oil % in F2. P2 (Sakha 3) and P3 (Sakha 4) were good combiner in F1 and F2 for all traits except plant height in F2. Seed yield of 10 crosses are significantly in F1, whereas in F2 Sakha 4 x S 3/25 is less than zero. No. capsules plant-1 the all crosses significantly greater than zero in F1, whereas in F2 two crosses less than zero (Sakha1 x Sakha 4 and S 3/25 x Olin). However, oil % was significantly greater than zero in F1 and F2 for all crosses. For oil yield plant-1 in F1 two crosses were less than zero (Sakha1 x S 3/25 and Sakha1 x Olin) whereas Sakha 1 x Sakha 4 is less than zero in F2. In most significant positive SCA were usually inconsistent across either generation.