In many Midwestern rivers, elevated concentrations of herbicides occur during runoff events for 1-3 months following application. The highest or 'peak' herbicide concentration often occurs during one of these runoff events. Herbicide concentrations in rivers are affected by a number of factors, including herbicide use patterns within the associated basin. Changing agricultural practices, reductions in recommended and permitted herbicide applications, shifts to new herbicides, and greater environmental awareness in the agricultural community have resulted in changes to herbicide use patterns. In the Midwestern United States, alachlor use was much larger in 1989 than in 1995, while acetochlor was not used in 1989, and commonly used in 1995. Use of atrazine, cyanazine, and metolachlor was about the same in 1989 and 1995. Herbicide concentrations were measured in samples from 53 Midwestern rivers during the first major runoff event that occurred after herbicide application (postapplication) in 1989, 1990, 1994, and 1995. The median concentrations of atrazine, alachlor, cyanazine, metribuzin, metolachlor, propazine, and simazine all were significantly higher in 1989/90 than in 1994/95. The median acetochlor concentration was higher in 1995 than in 1994. Estimated daily yields for all herbicides and degradation products measured, with the exception of acetochlor, were higher in 1989/90 than in 1994/95. The differences in concentration and yield do not always parallel changes in herbicide use, suggesting that other changes in herbicide or crop management are affecting concentrations in Midwestern rivers during runoff events.In many Midwestern rivers, elevated concentrations of herbicides occur during runoff events for 1-3 months following application. The highest or `peak' herbicide concentration often occurs during one of these runoff events. Herbicide concentrations in rivers are affected by a number of factors, including herbicide use patterns within the associated basin. Changing agricultural practices, reductions in recommended and permitted herbicide applications, shifts to new herbicides, and greater environmental awareness in the agricultural community have resulted in changes to herbicide use patterns. In the Midwestern United States, alachlor use was much larger in 1989 than in 1995, while acetochlor was not used in 1989, and commonly used in 1995. Use of atrazine, cyanazine, and metolachlor was about the same in 1989 and 1995. Herbicide concentrations were measured in samples from 53 Midwestern rivers during the first major runoff event that occurred after herbicide application (postapplication) in 1989, 1990, 1994, and 1995. The median concentrations of atrazine, alachlor, cyanazine, metribuzin, metolachlor, propazine, and simazine all were significantly higher in 1989/90 than in 1994/95. The median acetochlor concentration was higher in 1995 than in 1994. Estimated daily yields for all herbicides and degradation products measured, with the exception of acetochlor, were higher in 1989/90 than in 1994/95. The differences in concentration and yield do not always parallel changes in herbicide use, suggesting that other changes in herbicide or crop management are affecting concentrations in Midwestern rivers during runoff events.