Regarding the primary cultivation of Amaranthus hypocundriacus var. Cim and var. Kharkofski in Iran, evaluation of the direct and indirect effect of each yield component on the final yield, by the means of path analysis is highly important. This research was conducted in 2010 at the Veterinary Science Institute, Karaj, Iran. Experimental design was split plot in time in the form of a randomized complete block design with four replications. The main factor was planting season (spring and summer) and the sub factor was two amaranth varieties (Amaranthus hypocundriacus L. var. Cim and var. Kharkofski). Results indicated a significant positive correlation between stem yield with plant height (0.98**), leaf length (0.89**), leaf width (0.45*), petiole length (0.83**) and stem diameter (0.92**). There was also a significant correlation between total biomass yield with plant height (0.98**), leaf length (0.87**), leaf width (0.47*), petiole length (0.92**) and stem diameter (0.91**). The result of principle component analysis indicated that in the first prin, traits such as plant height, stem yield, leaf yield, flower and biomass yield contributed to about 80% of variations. The result of stepwise analysis of the traits that affect the dependent variable (biomass yield) indicated that four traits including flower yield, stem yield, leaf yield and petiole yield entered to the model respectively. The result of path analysis showed that stem yield had the highest positive direct effect on biomass yield and had determination of 0.482 of the total variations. Flower yield which was the first trait entering the model, was the second most effective trait on biomass yield with determination of 0.294. Therefore, it can be concluded that stem yield had the highest effect on biomass yield and after that, flower yield, leaf yield and petiole yield were the most effective traits on biomass yield, respectively.