Field experiment was conducted at the farm AL Sofia place in Ramadi /Anbar Province (22.35º north, 12.34º south) on the right side of the Euphrates during summer season of 2013. Arandomized complete block design with three replication. the aim of this study was to study genetic parameters and path analysis in soybean . in a split plot design three plant densities which were (80000 plant.ha-1) (66666 plant.ha-1) and (57142 plant.ha-1) as main plots and five soybean cultivars (Giza 35, Giza 111, Giza22 , Taka 2, ALshaima ).Estimated Genotypic, phenotypic variances and coefficient of variability, broad sence heritability percentage, genotypic and phenotypic correlation coefficient and other characters were estimated. Path coefficient analysis carried out to partition the genotypic correlation coefficient to the direct and Indirect effects.
The results showed that values variances and coefficients variation of genetic and phenotypic differed depending on plant density, and the values the genetic variation is greater than the values of variance environmental for all traits with the exception of the adjectives number of seeds in pod and weight of 100 seed, in the density D1 The values of the coefficient of variation phenotypic higher than the values of the coefficient of variation genotype for all traits. The highest heritability percentage in the broad sense of the trait number of seeds in pod amounted (96%).Explained analysis of the path coefficient that the number of seeds in pod for all plant densities achieved the highest direct effect in the yield of seeds was (1.156), while given the weight of 100 seed highest sum total of the effects of direct and indirect yield the seeds of (3.562) and all plant densities, we conclude from the study to adopt number of seeds in pod as a criterion for the development of electoral yield in soybean seeds.