When almost no broadleaf weeds were present, 0.56 kg/ha of the butoxyethanol ester and 0.84 kg/ha of the dimethylamine salt of 4-(2,4-dichlorophenoxy)butyric acid (2, 4-DB) caused no yield reductions of seedling alfalfa (Medicago sativa L.); but 0.56 kg/ha of 3,5-dibromo-4-hydroxybenzonitrile (bromoxynil) caused slight reductions. Rates of bromoxynil above 0.56 kg/ha caused significant yield reductions of alfalfa. Broadleaf weeds, when present, shielded alfalfa from the bromoxynil so that reduction in yields of broadleaf weeds and increases in alfalfa were similar to those obtained with 2,4-DB. At equivalent rates 3,5-dichloro-4-hydroxybenzonitrile (hereinafter referred to as chloroxynil) was less effective than bromoxynil for controlling broadleaf weeds and also was less injurious to alfalfa. When no weeds were present at time of application, bromoxynil at a rate of 0.56 kg/ha or greater and chloroxynil at 0.84 and 1.12 kg/ha significantly reduced yields of red clover (Trifolium pratense L.). When weeds were present bromoxynil did not reduce yields of red clover until the 1.12-kg rate was reached. Chloroxynil did not reduce yields at rates as high as 1.12 kg/ha. Excellent control of common lambsquarter (Chenopodium album L.) and common ragweed (Ambrosia artemisiifolia L.) was obtained, and no reduction in alfalfa or red clover occurred with 0.28 kg/ha of bromoxynil or chloroxynil mixed with 0.56 kg/ha of 2,4-DB ester.