在秋作小米(Setaria italica Beauv.)水田中，採用逢機完全區集設計分別進行11種萌前殺草劑及3種萌後殺草劑對其發芽率及植株生育影響之調查，並從事篩選試驗，以不間苗不除草(CK1)及間苗人工除草(CK2)為二個對照處理。結果發現萌前殺草劑以理有龍(50%WP, 1000X, 1kg/ha)及免草繁(21%WP, 4000X, 0.4kg/ha)效果較佳，對小米發芽率影響最小，分別為71.7%及76.6%。其在小米生育初期可有效抑制大部分之闊葉雜草，產量分別為1,810公斤/公頃及1,850公斤/公頃，惟顯著低於間苗除草之對照組，而且與不間苗不除草對照處理差異不顯著。萌後殺草劑則以本達隆(44.1%S, 200X, 31/ha)在小米分蘗期噴施效果明顯，對植株生育及產量影響最小，產量為1,978公斤/公頃，顯著高於CK1。

Eleven pre-emergence and three post-emergence herbicideswere screened in paddy field to investigate their ef1'ects on thegrowth of foxtail millet cropped ín the fall 1989. Two checkswere: treatment without thinning and weeding as check 1, andthat with thinning and hand weeding as check 2‘As the primaryresult we found that applicable pre-emergence herbicideswere linuron (50%WP, 1000X, 1 1jha)and sirius (21 %WP,4000X, O.4kgjha), and the feasible post-emergence was bentazon(44.1 %S, 200X, 31jha) which can be applied at tillering stageof millet. These two pre-emergence herbicides lowered the seedemergence rate of millet to 71.7% and 76.6%, and gave 1,810kgjha and 1,850 kgjha in yield, respectively. Their weeding effectwere on most wide-leaf weeds at the initial stage ofgrowth. Such yields were a little more than the check 1, butsignificantly less than the ckeck 2. Applying bentazon gave 戶heyield of 1,978kgjha, which was significantly higher than thecheck 1, but lower than the check 2.