In field experiments conducted on direct seeded wet land rice during Kharif 1986 and 1987, pre-emergence application of butachlor and pre-tilachlor/ fenclorim approved excellent control of grasses, good control of sedges and slight control of broad leaved weeds. On the other hand a mixture of thiobencarb and 2,4-D showed good control of grasses and excellent control of broad leaves but was -least effective against sedges. A mixture of thiobencarb with naproanilide obtained only a slight control of all the three classes. Bensulfuron-methyl, quinchlorac alone and a mixture of piperophos with 2,4-D satisfactorily performed against grasses but remained moderate against sedges. Pendimethalin and oxyfluorfen could not satisfactorily control all the three classes of weeds. None of herbicides under evaluation except quinchlorac caused any visual phytotoxicity to rice plants. Highest grain yield was recorded from hand weeded plots in 1986 and from butachlor followed by hand weeding treatment in 1987. All the treatments involving butachlor gave grain yields which were not significantly different from hand weeding treatment.