ABSTRACT
Sixtyfour Indian barnyard millet genetic resources were studied to assess the extent of genetic divergence for 18 quantitative traits using Mahalanobis D2 Statistic. Sixtyfour genetic resources of Indian barnyard millet were evaluated during kharif, 2020 in a lattice square design, replicated twice at Regional Agricultural Research. Based on Tocher’s method, the entire genetic materials were separated into eight distinct and non-overlapping clusters suggesting availability of genetic diversity. Cluster I has highest number of accessions (57) and the remaining were mono genotypic clusters II, III, IV, V, VI, VII and VIII containing only one accession each indicating high degree of heterogeneity among the genotypes. The widest inter-cluster distance was recorded between clusters IV and VII. Therefore, it is concluded that the genotypes to these clusters has to be intercrossed in order to create more diversity and improving grain yield in barnyard millet. The traits, grain yield per plant and panicle weight contributed maximum towards total divergence has scope of crop improvement through those traits. Diversity of parent paly main role for the successful of any breeding programme. So the crosses can be made among the parents having widest divergence would more likely to yield desirable recombination in segregating generations.