Two field experiments were carried out at in clay soil in Agricultural Research and Experimental Center of Agriculture at Faculty of Moshtohor, Benha University, Kalubia Governorate during 2009 & 2010 summer seasons to determine the critical period of weed competition between weeds and soybean crop through use of regression and economic approaches and also determine the relationship between weeds and soybean yield under three plant densities (105, 140 and 175 thousand plant/ fad.) and ten treatments  in two types of treatments (in the first type soybean was hand weeded for different periods 3, 6, 9 and 12 weeks after sowing (W.A.S.) and for whole season ,then no further weeding was done while, in the second type weeds were allowed to grow for different periods at 3, 6, 9 (W.A.S.) until harvest the weeds were removed by hand- weeded and weed competition treatment for whole season on seed yield of soybean and associated weeds.    The main results showed that, maximum yield losses of soybean due to weed competition in the whole season were 37.6 and 34.4 % from weed free treatments in 2009 and 2010 seasons. Mathematical model for the relationship between weed free and weed competition periods according to the recommended  losing yield value (10), clear that critical period of weed competition under 105 thousand plants density/fad., were (6.5 and 7.0) weeks for weed free as well as (4.25 and 4.25) for weed competition in 2009 and 2010 seasons, respectively. Under 140 thousand plant density the respective values were (6.2 and 6.3) weeks for weed free as well as (3.9 and 4.2) for weed competition  and for 175 thousand plant density were (5.9 and 5.6) weeks for weed free as well as (3.5 and4.15)  weeks for weed competition in 2009 and 2010 seasons, respectively . The critical period of weed treatments over plant densities were (6.45 and 6.4) weeks for weed free as well as (4.9 and 4.65) weeks for weed competition in 2009 and 2010 seasons, respectively. All weed competition exerted significant efficiency in controlling annual weeds. Weed free for the whole season treatment gave the best control for annual weeds gave the highest values of yield and yield components in two seasons. The sowing 175000 plant/fad. gave the lowest weight for dry weight for annual weeds and the tallest plants in first and second seasons. 105000 plant/fad., gave the best values of No. of branches and seed weight/plant in the first and second seasons. 140000 plant/fad., gave the highest value of No. of seed pod, weight of 100 seed and yield ton/fad. in the two seasons. Plots weeded at 3 and 6 week after sowing showed the best performance in all aspects of soybean. Such knowledge should be disseminated to farmers to keep soybean yield losses of weed competition to maintain maximum soybean seed yield. This study showed that negative correlation for annual weed weight and all studied characters under study.