Ten flax genotypes (G) were evaluated over six environments (E) “combination of three years (2013/14, 2014/15 and 2015/16) and two locations (Giza Exp. Sta., Giza Governorate and Ismaelia Exp. Sta., Ismaelia Governorate) in Egypt to determine genotype X environment (GE) interaction and stability. Mean squares of genotypes and Environments (E) were highly significant for all studied characters, indicating that the genotypes differ in their genetic potential as well as variability among the environments studied. Also, GxE interaction was significant for all characters except, straw weight per plant, 1000-seed weight and no of seeds per capsule. This result indicated that genotypes had considerable different responses to environmental conditions. The significant variance due to residual (pooled deviation) for all characters indicated that genotypes differed with respect to their stability suggesting that prediction would be difficult, which means that mean performance alone (mean yield) would not be appropriate. Estimates of variance components, phenotypic (PCV) and genotypic (GCV) coefficients of variability and broad sense heritability (H) indicated the possibility of using plant height as selection index for improving straw weight per plant and both of 1000-seed weight and no. of capsules per plant in selection index for improving seed weight per plant. Based on estimates of mean performance (`x ) and four stability parameters (regression coefficient (b), deviation from regression (S²d), coefficient of determination (r²) and the ecovalence stability index (w)), it could be concluded that L.541-D/5 could be considered as ideal genotypes for all studied characters as well as L.541-C/8 could be considered as ideal genotypes for the three traits which related of seed (seed yield per fed, oil yield per fed and oil percentage). Thus, they are recommended to be released as stable high-yielding cultivars and/or to be incorporated in the breeding stocks in any breeding program aiming to produce stable genotypes for the above-mentioned characters. Phenotypic correlation coefficients among straw weight and other components indicated the possibility of selecting genotypes characterized by high straw yielding ability and in the same time high seed yield potentialities. However, seed yield per plant was significant positively correlated with both no. of capsules per plant and 1000-seed weight. Also, no. of capsules per plant exhibited highly significant positive correlation with 1000-seed weight. These results indicate the possibility of using plant height as a selection index for improving straw weight per plant and both of 1000-seed weight and no. of capsules as selection indices for improving seed weight per plant.