This study was conducted to evaluate the effect of hairy vetch cover crop residue on runoff losses of atrazine and metolachlor under both no-till corn field plots and from a laboratory runoff system. A 2-year field study was conducted in which losses of atrazine and metolachlor from vetch and non-vetch field plots were determined from the first runoff event after application (5 and 25 days after application in 1997 and 1998, respectively). A laboratory study was conducted using soil chambers, designed to simulate field soil, water, vegetation, and herbicide treatment conditions, subjected to simulated rain events of 5, 6, 20 and 21 days after application, similar to the rainfall pattern observed in the field study. Atrazine losses ranged from 1.2 to 7.2% and 0.01 to 0.08% and metolachlor losses ranged from 0.7 to 3.1% and 0.01 to 0.1% of the amount applied for the 1997 and 1998 runoff events, respectively. In the laboratory study, atrazine runoff losses ranged from 6.7 to 22.7% and 4.2 to 8.5% and metolachlor losses ranged from 3.6 to 9.8% and 1.1 to 4.7% of the amount applied for the 5-6 and 20-21 day events, respectively. The lower losses from the field study were due to smaller rainfall amounts and a series of small rains prior to the runoff event that likely washed herbicides off crop residue and into soil where adsorption could occur. Runoff losses of both herbicides were slightly higher from non-vetch than vetch field plots. Losses from the laboratory study were related to runoff volume rather than vegetation type.