Multi-environment test of genotypes for yield performance always subjected to genotype environment interaction, which complicates selection of stable genotypes. Sixteen cowpea genotypes (fourteen promising genotypes and two released varieties) were arranged in 4x4 lattice design with three replicates and evaluated at six environments for grain yield performance in 2016 main cropping season. AMMI combined analysis of variance revealed that genotype, environment and GEI significantly affected (P≤0.01) cowpea grain yield. Both IPCA1 and IPCA2 were highly significant (P≤0.01) and captured, 67.96% and 20.04% of the variability relating to GEI, respectively. AMMI biplot revealed, G1, G2, G9, G13, G14 and the two released varieties as stable or broadly adapted genotypes. Based on ASV rank, G1, G14, G2, G9, G5, G15, and G13 scored lower ASV value and were the most stable genotypes, whereas G10, G3, and G4 which scored higher ASV value were the most unstable. According to GSI rank, G14, G2, G5, G1, G9, G15 and G16 were both high yielder and stable genotypes. GGE biplot identified three distant mega environments and their specifically adapted genotypes which suggests the possible existence of different environments. Considering all stability analysis, G1 (OT-03L-2046-2), G2 (IT93K-2046-1), G5 (IT-97K-569-9), G9 (IT-97K-356-1), and G14 ((IT-95K-268-1-4) were the most stable among genotypes. Therefore, after repeating of the study over seasons and across locations, they could be recommended and verified for release.

Keywords: Multi-environment trial, Vigna unguiculata, GEI, Broad adaptation