Transport of pesticide to groundwater is governed in part by sorption of the pesticide to soil particles. Sorption may be dependent on soil moisture conditions, but limited data are available from which to elucidate the effect. The objective of this research is to determine the effects of soil water content and matric potential on the sorption coefficient of a hydrophilic pesticide using an unsaturated flow method. Sorption of dicamba was determined in three soils, each at two water contents. At low water contents (~0.05 kg kg^-1), sorption coefficients were low and similar for all three soils, ranging from 0.02 L kg^-1 for the sandy loam to 0.06 L kg^-1 for the silty clay loam. At higher water contents (0.20 to 0.25 kg kg^-1), the sorption coefficient for the sandy loam was unchanged, for the silt loam was doubled, and for the silty clay loam was increased by almost tenfold. Matric potential per se did not affect the sorption coefficients, as they were similar for all three soils at the low water content while matric potential varied by five orders of magnitude. These data are consistent with the hypothesis that access to sorption sites decreases as water content decreases. Furthermore, the data suggest that sorption is more influenced by water content in soils with a greater number of sorption sites. The effects of water content on pesticide sorption require further research and may ultimately have implications for the methods used to determine sorption and for managing pesticide application.