Sainfion ( onobrychis viciifolia Scop) is a perennial legume widely adapted to environmental conditions and has been successfully used as a pasture and hay forage in Iran . This experiment was carried out to investigate adaptability and genetic variation of different traits using 21 Iranian Sainfoin ecotypes during two years. The experiment was conducted using a randomized complete block design with three replications. Results indicated that there are considerable within and among population variation in the studied germplasm, can be exploited in breeding programs. Genotypic coefficient of variability for forage yield was highest while the lowest belonged to phenological traits and plant height. The heritability for dry matter yield, leaf percentage, stem percentage, plant height and number of stem per plant were over 50 percent. The result of analysis of variance showed significant differences among ecotypes for all traits. The interactions of year by ecotypes for all characters were highly significant showing different response of ecotypes. Results showed that one ecotype from Khansar had the highest value for dry matter production and the lowest one belonged to an ecotype from Semirom. Evaluation of traits related to forage quality revealed that the highest proportion of leaf to stem ratio belonged to ecotypes from Faradombe and Khansar (2.07 and 2.04) respectively. Correlation coefficients of dry matter yield with fresh yield, dry matter percentage, stem percentage, number of stem per m 2 , and number of node were significantly positive. Also correlation coefficients of dry matter yield with leaf percentage and leaf stem ratio were significantly negative. On the basis of stepwise regression analysis, stem percentage and plant height justified the majority of dry matter yield variation. The result of path analysis indicated that stem percentage and plant height had the most direct and positive effect on dry matter yield. Factor analysis revealed five factors which justified more than 75 percent of the total variation and were named forage yield components, viability potential, single plant size, establishment rate and dry matter percentage, respectively. Applying cluster analysis on the studied attributes, three groups were obtained. Regarding all measured characters the least and the most similar accessions were identified to be used in the future breeding programes. Key Words: Cluster analysis, Correlation coefficients, Factor analysis, Genetic diversity, Sainfoin.
