Sorption and desorption of metribuzin [4-amino-6-(1,1-dimethylethyl)-3-(methylthio)-1,2,4-triazin-5(4H)-one] were measured in Dundee silty clay loam surface and subsurface soils. Soil samples were taken from a profile that had been divided into six sections from the surface to a depth of 175 cm. Metribuzin sorbed weakly to all soils from this profile with Freundlich distribution constants ranging from 0.78 to 1.34 μmole/kg. Soils from lower depths of the profile having higher pH (>7.0) and clay contents (>35%) sorbed significantly more metribuzin than the lower clay content, higher organic matter surface soils. Stepwise regression of the distribution constants against the variables pH, organic matter, clay content, and sand content showed that clay was the single best predictor, with sorption increasing as clay content increased (r2 =0.750). The combination of two variables most related to sorption was clay and pH (r2 =0.860, P= 0.15). Organic matter was not one of the primary variables related to sorption. Metribuzin was easily desorbed in all soils with less than 5% of the originally applied metribuzin remaining after three desorption treatments. This would indicate little tendency for irreversible sorption. This study demonstrates that soil properties within a profile determine sorption and, subsequently, movement of metribuzin.