Fifty genotypes of finger millet were studied for genetic divergence using D 2 analysis for ten quantitative traits. These genotypes were grouped into fourteen clusters. Out of which, Cluster I had seventeen genotypes, cluster III ten genotypes, cluster V six genotypes, cluster VI seven genotypes, cluster II, cluster IV and clusters VII to XIV consisted of one genotype each. Considering the inter cluster distances, it was the highest between cluster XI and XIII followed by the inter cluster distance of V and XI and XIII and XIV. Cluster XIV (genotype MS 3221) had the highest mean for grain yield and productive tiller number. For number of fingers, cluster IV (genotype TNAU, 972) and cluster XIII (genotype Indaf 11), for finger length cluster X (PR 202) and for thousand grain weight cluster XI (AF 260) were the superior clusters for the respective characters, the genotypes of which can be effectively used in breeding programmes. Among the characters, days to 50% flowering followed by number of leaves contributed more towards the total divergence. INTRODUCTION Finger millet {Eleusine coracana (L.) Gaertn} is cultivated under varied ecological conditions for both grain and fodder purposes in Africa and South Asia. This crop has accumulated considerable diversity over the years for vegetative, reproductive and physiological characters. The nature and magnitude of genetic divergence play an important role in the formulation of successful breeding programme. The genetically diverse parents are needed to produce heterotic effect and desirable segregants. The D 2 statistic analysis is a powerful tool in quantifying the degree of divergence among the populations. Thus, the present study was undertaken to know the nature and magnitude of genetic diversity in finger millet and to identify genetically divergent genotypes to serve as parents in hybridization programme.