Quantitative measurements were made by cucumber root bioassay of 2,4-dichlorophenoxyacetic acid (2,4-D) contained in washoff (water-soil mixture) from cultivated fallow Cecil sandy loam soil. Formulations of 2,4-D used included iso-octyl and propylene glycol butyl ether esters and an alkanolamine salt of the ethanol and isopropanol series. Simulated rainfall intensities and storm durations used represented storm frequencies of 1, 10, 80, and greater than 100 years. The 2,4-D applied at rates of 2.2 or 4.4 lb/A was assayed in washoff and as residue in the soil. Concentrations of 2,4-D in washoff were positively correlated with the rate applied, were greatest early in each storm, and decreased with duration of the storm. The iso-octyl and butyl ether ester formulations of 2,4-D were far more susceptible to removal in washoff than the amine salt. When the amine salt was used, 2,4-D concentrations in washoff were less than 1 ppm, even in the early stages of runoff, whereas concentrations as high as 4.2 ppm were measured using an iso-octyl ester formulation of 2,4-D. When a direct comparison was made between the butyl ether ester and amine salt forms of 2,4-D, ester or amine losses were respectively, 13 and 4% following a 1-year frequency storm and 26 and 5% following a 100-year frequency storm. Soil bioassays showed that most of the 2,4-D remained in the surface 3 in of soil.