In order to have a healthy environment, high productivity and reduced costs, it is essential to use lower rates of herbicides in chemical weed control. To investigate the effects of reduced rates of herbicides on weed control and soybean yield, an experiment was carried out in a randomized complete design with 11 treatments and four replications at experimental station of Agriculture College of Sari, Iran, during the growing season of 2000. Three herbicides were used at three rates each, namely, the recommended rate, 50 and 75% of the recommended rate of trit10uralin (pre - plant incorporation), bentazon, and sethoxydim (post emergence). Eleven treatments were triflouralin (pre - plant incorporation) at 0.84 kgailha (recommended rate), 0.63 and 0.42 kgai/ha, sethoxydim (post emergence) at 0.224 kgai/ha (recommended rate), 0.168, 0.112 kgai/ha and bentazon (post emergence) at 0.84 kgai/ha (recommended rate), 0.63, and 0.42 kgai/ha and two control treatments of weed - free check and weedy - check. Considering soybean seed yield of weed free check, reduction of herbicide application is possible to 50 percent of recommended rate. All treatments could control velvetleaf (Abutilon teophrasti) and pigweed (Amaranthus retroflexus) more than ninety percent (90%). The best grass control was seen in 50% of the recommended rates of trit10uralin and sethoxydim. The herbicides had no adverse effect on soybean. With reduced herbicide rates, both weed control percentage and soybean yield did not show any significant difference in comparison to recommended rate.