S. Bhujel, N. R. Adhikari , A. Shrestha, S. Poudel Variability, correlation, diversity, yield, rice 38-46 5 (2) 2021 April,2021 The present experiment was conducted at Regional Agricultural Research Station, Parwanipur, Bara, Nepal during 2019 with 44 rainfed rice genotypes including four released varieties in Randomized Complete Block Design (RCBD) with three replications. The objective of the study was to assess the genetic variability in the rice genotypes to identify the promising rice genotypes with high yield potentials under rainfed situation. Wide ranges of variation among the forty-four genotypes were recorded. Genotype HHZ6-DT1-L11-L11 has been observed high grain yielder (6.792 ton per hectare). Correlation analysis showed that number of filled grains per panicle, dry matter content and total number of grains per panicle showed positively significant correlation with grain yield whereas days to 50% flowering, days to maturity and number of tillers per m2 showed negatively significant correlation with grain yield. Path analysis using grain yield as a dependent variable revealed that days to 50% flowering, test weight, number of filled grains per panicle, dry matter and number of grains per panicle had positive direct effects with grain yield. The dendogram resulting from UPGMA analysis of forty-four genotypes formed six clusters, where Cluster 1, Cluster 2, Cluster 3, Cluster 4, Cluster 5 and Cluster 6 represent 47.2%, 9.09%, 6.81%, 29.4%, 2.32% and 4.64% respectively. Genotypes under Cluster-II was higher yielder and genotypes under cluster-III was early flowered and early matured. Hence, genetic diversity has been observed among the tested genotypes so that direct selection criteria for higher grain yield is possible by evaluating the days to 50% flowering, days to maturity, plant height, number of tillers per m², number of filled grains per panicle, and thousand grain weight.