Leaching and runoff losses of the postemergence-applied herbicide dicamba were evaluated over a 3-yr period (1989 to 1991). Dicamba was applied at the recommended rate (0.56 kg ai/ha) to conventional and mulch tillage planted corn fields on Hagerstown silty clay loam (fine, mixed, mesic Typic Hapludalf). Mulch tillage followed several years of no-tillage corn. Root zone leachates were collected utilizing pan lysimeters placed 1.2 m below the soil surface. Surface runoff was monitored and collected with an HS-flume and automated sampling equipment. Leaching was greatest during 1989, and runoff events were recorded only during this season. Leachate samples containing measurable levels of dicamba were obtained within 21 d of herbicide application or within slightly more than one soil half-life of this chemical. More dicamba leached under mulch tillage than conventional tillage management. Tillage rotation (no tillage to mulch tillage) did not alter the leaching loss potential of dicamba beneath the minimally tilled soil surface, as postulated, compared with the previous untilled surface. The mulch tillage surface reduced runoff water losses compared with conventional tillage, but early-season leaching activity, coupled with the minimal persistence of dicamba in soil, negated runoff transport of this herbicide from either tillage system when the first runoff event occurred 12 d after its application.