Two field experiments were carried out from 1993 to 1995 to evaluate the critical period of weed control and to develop suitable weed management practices for jungle rice, horse purslane, and cockscomb in soybean. Horse purslane was more competitive during early growth stages (up to 45 days after sowing [DAS]) and cockscomb was more competitive during later growth stages, whereas jungle rice was competitive throughout the growing season. The critical period of weed control was found to be 30 to 45 DAS. Weed-free maintenance up to 45 DAS resulted in a 74% increase in grain yield of soybean over the unweeded control. Keeping soybean weed free for 45 d or allowing weeds to remain in the crop for less than 30 d resulted in no significant yield loss. Sequential application of a reduced rate of soil-applied trifluralin 1.0 kg ha–1 (0.67 ×) with postemergence fluazifop 0.75 kg ha–1 (0.75 ×) or a reduced rate of soil-applied trifluralin or pendimethalin at 1.0 kg ha–1 (0.67 ×) followed by hand hoeing 35 DAS provided better control of a broad spectrum of weeds than a single application of a postemergence herbicide applied at reduced or recommended rates. Integration of reduced rates of soil-applied herbicides with post-emergence herbicides or hand hoeing 35 DAS produced soybean yields similar to the hand-weeded treatment. Compared to the weed-free or integrated weed control, a single application of soil-applied or postemergence herbicide did not control a broad spectrum of weeds and reduced soybean yield.