Soil factors that may affect activity of the propynyl ester of CGA-82725 {2-[(3,5-dichloro-2-pyridinyl) oxy] phenoxy] propanic acid}, methyl ester of haloxyfop {2-[4-[[3-chloro-5-(trifluoromethyl)-2-pyridinyl] oxy] phenoxy] propanoic acid}, and butyl ester of fluazifop {(±)-2-[4-[[5-(trifluoromethyl)-2-pyridinyl] oxy] phenoxy] propanoic acid} were examined. Adsorption of the three herbicides on three Illinois soils was similar but very low. Linear partion coefficients (Kd) ranged from 0 to 1.75 for CGA-82725 and haloxyfop, while no adsorption was detected for fluazifop on Drummer silty clay loam, Cisne silt loam, or Bloomfield sand. Conversion to the parent acid (active form) following application to the soil was rapid. The acid form was predominant for 2 months. Two metabolites were recovered; one was common to all three herbicides and intermediate in polarity to the acid and ester forms, and the other, unique to CGA-82725, was slightly more polar than the acid form. Extractability of the herbicides decreased dramatically over time. Haloxyfop and CGA-82725 leached to similar depths in soil column studies. Fluazifop tended to move more readily in the silty clay loam but less rapidly in the sand. Adsorption studies with 15 additional soil or sediment samples with widely varying physiochemical properties yielded adsorption (Kd) values of 0.79, 1.09, and 1.15 for fluazifop, haloxyfop, and CGA-82725, respectively, when averaged over all samples studied. Adsorption did not appear to be significantly influenced by organic carbon content, clay content, or pH.