In India seventy six landraces of rice were characterized based on quantitative descriptors along with grain characters to study genetic diversity employing D2 analysis. These descriptors are highly heritable, unambiguous and easily identifiable. Based on D2 analysis the genotypes were grouped into six clusters. Dendrogram showed that out of six clusters, cluster I was the largest comprising of 49 landraces, followed by cluster III with 14, cluster II with 7 and cluster IV with 4 and remaining two clusters V and VI, were monogenotypic clusters. Maximum intra cluster distance was observed in cluster II, followed by cluster III, cluster I and cluster IV, indicating that some genetic divergence still existed among the genotypes within each of these clusters. Maximum inter cluster distance was exhibited between clusters II and VI followed by clusters I and VI, clusters V and VI and clusters II and IV. The greater the distance between two clusters, the wider the genetic diversity among the genotypes of those clusters. Such highly divergent, high performing genotypes could be used as parents for future recombination breeding programme in order to get high heterotic recombinants. Out of the seventeen characters studied, six characters namely stem length, grain zinc concentration, gel consistency, amylose content, spikelet fertility and grain iron concentration together contributed maximum (97.33 %) to the total genetic divergence which should be given importance in further breeding programmes.