Sorghum (Sorghum bicolor (L.) Moench) is an important essential cereal crop in Ethiopia. Conversely, its productivity is low due to numerous biotic and abiotic factors. There are diverse and dynamic environmental conditions which needs detail and continue study on genotypes by environment interaction (GEI) to develop stable genotypes. The objective of this study was to determine the magnitude of GEI for grain yield of forty two sorghum genotypes and to identify stable and high yielding genotypes across locations. The experiments were laid out at three locations for two growing seasons using alpha lattice design with three replications. The plot size 5 m x 0.75 m x 2 rows (7.5 m<sup>2</sup>) and distance between block, replication, and plot was 1m, 1.5m, and 0.75m, respectively. Phonologic, agronomic, diseases and grain yield data were collected but only grain yield was used for stability analysis. The ANOVA revealed highly significant variation (p <0.01) among sorghum genotypes across locations and seasons. Mean grain yield of genotypes ranged from 1.29 to 3.69 with mean grain yield of 2.36, while environment range from 1.18 to 3.63 t/ha. The genotype G1 showed good performance across all test sites which range 5<sup>th</sup> at E1,3<sup>rd</sup> at E3 and E4, 15<sup>th</sup> and 7<sup>th</sup> at E5 and E6 and maximum grain yield was harvested from E3. Yield data were also analyzed using the GGE (that is, G, genotype +GEI, genotypes-by- environment interaction) bi-plot method. The first two principal components (PC1 and PC2) were used to create a 2- dimensional GGE bi-plot and explained 59.67 and 13.48 % of GGE sum of squares, respectively. GGE bi- plot identified G16, G4, and G1 high yielders and stable and G34 and G25 was the lowest yielding and least stable across locations. On the other hand, the environment E6, E4 and E1 were the most suitable to select desirable genotypes.