Fluazifop-butyl {(±) butyl 2-[4-[[5-(trifluoromethyl)-2-pyridinyl] oxy] phenoxy] propanoic acid}, haloxyfop-methyl {methyl-2-[4-[[3-chloro-5-(trifluoromethyl-2-pyridinyl] oxy] phenoxy] propanoic acid}, and sethoxydim {2-[1-(ethoxyimino)-butyl]-5-[2-ethylthio)propyl-3-hydroxy-2-cyclohexen-1-one}exhibited soil herbicidal activity to annual grass species in greenhouse and field experiments. Control of forage sorghum [Sorghum bicolor (L.) Moench. ‘Rox Orange’] with all three herbicides was greater when seeds were planted at or near the soil surface than when planted 2, 4, or 6 cm deep. Imbibition of the herbicides by forage sorghum and soybean [Glycine max (L.) Merr.] seeds did not reduce germination at concentrations of 10-3 M or lower, but forage sorghum seedling survival and vigor were reduced by herbicide concentrations as low as 10-7 M. Fluazifop-butyl, haloxyfop-methyl, and sethoxydim were more phytotoxic when applied to the root zone than the shoot zone. Haloxyfop-methyl exhibited the longest soil residual control in the field followed by fluazifop-butyl and sethoxydim. Haloxyfop-methyl and fluazifop-butyl controlled simulated shattercane [Sorghum bicolor (L.) Moench. ♯3 SORVU] in soybeans with preemergence and early postemergence applications, whereas control from similar applications of sethoxydim was limited and variable.