Rice (Oryza sativa L.) is the staple food for more than 50 per cent world's population and 85 per cent Indian population. The rice production areas in the country are very diverse; hence, evaluation of genotype for its stable performance across the environment is very important. The study was undertaken to identify the stable rice variety in demonstration trials conducted on eleven varieties tested in five environments in rabi season. Among many available statistical techniques, the additive main effect and multiplicative interaction (AMMI) has been extensively used for GE interaction and stability analysis. First two PCA's are significant at 1 per cent and explained 89 per cent of the total variation with 24 degrees of freedom. The biplot obtained from AMMI analysis displayed PCA scores plotted against each other provides visual inspection. AMMI Stability Index (ASI) has been used to get quantitative value for stability analysis which helps in interpretation of the results. But, stable genotype may not always be high yielder which is the major requirement of farmers; therefore a new index namely 'Rank Based Stability Index (RBSI)' is proposed which identifies the stable genotype with high yield. The result showed that G1 and G2 were the most stable varieties with high grain yield, whereas G7 and G11 were found to be least stable. All eleven varieties were grouped into four clusters. First cluster consists of two varieties G1 and G2; second cluster contains four varieties G6, G10, G5 and G9; third cluster includes two varieties G4 and G11; and fourth cluster comprised of three varieties G8, G3 and G7.