-

This research was conducted in order to have genetic evaluation of some important agronomic traits as related to seed yield, using 14 soybean varieties from different maturity groups in a randomized complete block design with four replications at Bayo Cola Experiment Station of Mazandaran Research Center in 1991 growing season. The traits under evaluation were seed yield (gr/plot), plant height (em), number of pods/plant, number of seeds/plant, number of seeds,! pod, number of nodes/ plant, pod length (em), number of branches/plant, lOa-seed weight (gr), number of days to flowering, harvest index, length of internode (em), number of pods/node, and seed filling period ( days). Analysis of variance showed that there were significant differences among varieties in terms of traits under study, indicating the existance of genetic variation among varieties. Among the desired traits, number of pods/plant, number of seeds/plant, and IOO-seed weight. al] of which are considered yield components, had the highest amount of correlation with seed yield. Using multiple linear regression model. 12 traits under study, showed 89% of total variations in seed yield, were attributed to four traits; namely number of pods/plant, number of seeds/plant, IOO-seed weight, and number of seeds/pod while 89% was attributed to all the traits under study. Results of path analysis showed that the highest direct effect, being positive, was related to number of seeds/plant and the lowest direct effect, which was negative, was related to the number of seeds/pod. Number of seeds/plant in the first place, number of pods/plant and 100-seed weight in the second and third places, each or all together. can be used directly or indirectly as selection criteria for identification of high yielding genotypes in segregating generations. Results of Factor analysis showed the importance of those four traits through an important factor, called yield component. This factor explained 21 % of total variation in data.