Abstract: Five soybean genotypes (Giza 35, Crawford, Giza 82, Clark and Giza 111) were evaluated in six sowing dates (Feb. 25, Mar. 25, Apr. 25, May 25, June 25 and July 25) during the two consecutive growing summer seasons of 2000 and 2001 at Agricultural Experimental Research Station in Dirab, Saudi Arabia to explore the genotypes x environment effects and stability in performance of soybean genotypes for seed yield. Genotypes, years, sowing dates and their interaction were highly significant for the studied traits. The significant G x E interaction for seed weight per plant (g) and seed yield (t ha-1) indicated that the tested genotypes ranked differently across diverse environments. The response of seed weight and seed yield varied from genotype to another across different environmental conditions. This may offer raw material for improving soybean performance under the investigated conditions. Giza 111 and Clark had high means performance and had high phenotypic stability and they could be grown over wide range of environments. Results showed that Giza 111 and Clark have the highest mean values, while Giza 82 and Giza 35 had the lowest mean values over all environments and poorly adapted. The results suggested that to maximize seed yield potential, genotypes which have a consistently high yield performance across diverse growing environments should be selected and more than one year of evaluation. Genotype x environment interactions proved to play a significant role in the success of any breeding programs for development of genetic materials adapted to wide range of environments.