In the flood‐prone area, the improved high‐yielding variety of rice declines its vegetative growth and suffers substantial yield losses due to water stagnation. To address the issue measurement, the present work implemented 45 rice genotypes in a randomized complete block design with three replicates over three consecutive years (2018–2021) across submergence (E1, E3, E5) and optimum environments (E2, E4, E6) to assess the adaptability of these genotypes and identify the most desirable type by various stability indices. A significant genotype × environment interaction (GEI) was found in the combined ANOVA of yield and its component traits. The combined analyses of yield component traits through the multi‐trait stability index (MTSI) and yield‐stability statistics (YSi) found the most promising genotypes G20 and G32, respectively. The mean grain yield advantage was found in G18 under submerged conditions and G25 across the environments. A highly significant correlation (p < 0.01) exists between the stress tolerance index and yield in both submergence (r = 0.96) and normal (r = 0.90) circumstances. Among genotypes G18, G5 and G19, G20 showed strong stability for grain yield based on univariate stability parameters (YSi, σi2, Wi2, S2d and bi). An AMMI1 biplot analysis indicated that genotypes G37, G45, G32, G31, G27, G28, G19, G17 and G7 exhibited stability for grain yields, with IPCA1 values approaching zero. The GGE biplot analysis on yield was constructed into two mega‐environments, where G18 (Narkel Chari), G31 (CR Dhan‐500), G9 (Bhasha Manik) and G40 (SS‐1) were winners of submergence stress and G25 (Narayan Kamini) was the winner of normal environments.