The present study was conducted on eight barley genotypes with varying resistance to leaf blight of barley caused by Bipolaris sorokiniana across five environments. GGE biplot analysis was carried to study the relationships among test environments, discriminating ability of test environments and representativeness of test environments. Based on analysis, it is revealed that the five environments fell into three apparent groups: Faizabad, Karnal and Hisar formed one group, Varanasi formed another group and Pantnagar formed the third environment. Among the five environments, Karnal & Faizabad are most representative whereas Varanasi and Pantnagar areleast representative. Test environments that are both discriminating and rep resentative (Karnal & Faizabad) are good test environments for selecting generally adapted genotypes. Discriminating but non-representative test environments (Pantnagar) are useful for selecting specifically adapted genotypes if the target environments can be divided into mega-environments. They are useful for culling unstable genotypes if the target environment is a single mega-environment. Non-discriminating test environment i.e. Varanasi (those with very short vectors) is less useful because it provides little discriminating information about the genotypes.