|The present investigation was carried out at Seed Breeding Research Farm, Department of Plant Breeding and Genetics, Jawaharlal Nehru Krishi Vishwavidyalaya, Jabalpur (M.P.) during rabi season of 2010-11. Analysis of variance revealed significant differences among the genotypes for all the characters studied except for number of seed per capsule. The highest magnitude of genotypic variation was observed for 1000-SW, number of capsules per plant, oil content, number of seeds per capsules, number of capsules per plant number of primary branches per plant, plant height. High heritability coupled with high genetic advance was observed for dry matter of capsules, dry matter of leaves, plant height, biological yield, dry matter of branches, total dry matter of plant, dry matter of main stem and seed yield per plant indicating that heritability is due to additive gene action and selection based on these characters may be effective. Correlation coefficient analysis revealed that biological yield, number of capsules per plant, number of secondary branches per plant, number of primary branches per plant, dry matter of capsules, total dry matter of plant and 1000-seed weight indicated strong positive correlation with seed yield per plant at both genotypic and phenotypic level, whereas, dry matter of branches showed high genotypic correlation but moderate phenotypic correlation, while, oil content and plant height expressed moderate positive genotypic correlation. Path coefficient analysis revealed that dry matter of capsules, biological yield, harvest index, dry matter of branches, dry matter of main stem and dry matter of leaves, showed high positive direct effect on seed yield per plant, and selection for these characters would lead to increase in yield. Divergence analysis revealed that considerably low amount of genetic divergence was present in the material under study. Based on Mahalanobis’s D2 statistic, genotypes were grouped into three non-overlapping clusters. Intercrossing of genotypes from different clusters showing superior mean performance may help in obtaining higher yields.