The present study was conducted with 166 germplasm of linseed (Linum usitatissimum
) including three checks, to assess the extent and pattern of genetic divergence. The experiment was conducted in Augmented Design II (AD-II) at Sam Higginbottom University of Agriculture, Technology and Sciences Allahabad (U.P) India during rabi 2016-2017. The genetic diversity was estimated by Mahalanobis D2
statistics and the genotypes were grouped into 13 clusters using Tochers method each clusters having 15,3,11,20,4,11,14,17,16,19,15,9 and 12 genotypes respectively. The highest intra-cluster distance was found in cluster III followed by in descending order cluster II, IV, I and V, suggesting possibility of easy trait manipulation between these genotypes for linseed improvement. Cluster IV, XIII, VII and VI showing the highest mean for primary branches per plant number of capsule per plant, number of seeds per capsule and seed yield per plant, appears a desirable parent. Principle component analysis (PCA) showed the first four PCs had eigen value >1.00 and accounted more than 71.84% of total variation. The predominance nature of yield contributing traits like
number of primary branches per plant, number of capsule per plant and number of seeds per capsule, indicated that these components were proven more important and crossing between and/ or within cluster may give good recombinants for linseed improvement.