Plant breeders are very concerned about the presence of genotype-environment (G x E) interaction since it might make it more difficult to identify superior genotypes and limit the benefits of selection. In each season of 2019 and 2020, three separate experiments were conducted according to sowing dates. The trial design was laid out in a split-plot design with three replicates in each of the sowing dates and seasons. The densities were located in main plots, while, six soybean genotypes were grown in a sub-plot. To evaluate the plot yield performance, quantify (GxE) interaction and screen genotype using various stability techniques. Significant mean squares for genotypes, environments, and (GxE) interaction were found, revealing that genotypes responded differently to the various environmental factors. The terms of predictable (linear) and unexpected (non-linear) interaction components were highly significant further supports the fact that the tested soybean genotypes varied greatly in their relative stability. The most plot weight was generated by Giza 22 and H18 L54 that out-yielded the grand mean across studied environments. The genotype H18 L54 was stable phenotypically because, it had S²d_i values not significantly different from zero and bi = 1. Also, it was averagely stable under the studied environments. According to AMMI and GGE biplot analysis, The genotype H18 L54 was more stable as located nearest to the origin and with above-average mean descending ranked as follows: Giza22 > H18 L54, whereas the remaining genotypes had below-average mean yield. The mention genotypes seemed to be ideal across various environments.