The problem of genotype by environment interactions that often complicates the interpretation of multi-locations trail analysis making the prediction of genotype performance difficult can be eased with the adoption of the Additive Main Effects and Multiplication Interaction (AMMI) model analysis. The AMMI model was used to evaluate tuber yield stability of twelve potato (Solanumtuberosum L.) genotypes in randomized complete block design with three replications at Sinana, Shallo and Dinsho during 2009, 2010 and 2011. The objectives were to estimate the nature and magnitude of GEI for tuber yield and to identify stable potato genotypes for general adaptation and unstable genotypes for specific adaptation. Combined analysis of variance showed highly significant difference between the genotypes, locations and GEI. Proportion of variation captured by genotypes, locations and GEI is 13.65, 51.64 and 34.81, respectively indicating more effects of locations as compared to genotypes.