Field experiments were conducted during kharif (June-July), 2002, and late rabi (December-January), 2003 on sandy loam soils of Coimbatore to manage the late emerging weeds in irrigated groundnut. In screening trial, herbicides, metolachlor and fluchloralin were tested at four different levels viz., 0.75, 1.00, 1.25 and 1.50 kg/ha. The experimental fields were dominated by Setaria verticillata in grasses, Cyperus rotundus in sedges and Trianthema portulacastrum in broad-leaved weeds. Higher levels of herbicide dose recorded more weed control efficiency, yield and yield attributes of groundnut irrespective of the herbicides without any adverse effect on the crop. Among the levels, 1.00 kg/ha produced similar results with that of 1.25 or 1.50 kg/ha. Sequential applications of metolachlor as pre emergence at sowing followed by one on 40 DAS preceded with earthing up reduced the weed density, dry matter accumulation and nutrient removal by weeds significantly and comparable with weed free check. Initial reduction on the load of Rhizobium gained its original strength at later stages of crop growth.