In many watersheds, nonpoint-source contamination is a major contributor to water-quality problems. In response to the recognition of the importance of nonpoint sources, the Wisconsin Nonpoint Source Water Pollution Abatement Program (Nonpoint Program) was enacted in 1978. This report summarizes the results of a study to assess the effectiveness of watershed-management practices for controlling nonpoint-source contamination for the Eagle Creek and Joos Valley Creek Watersheds. Streamflow-gaging stations equipped for automated sample collection and continuous recording of stream stage were installed in July 1990 at Eagle and Joos Valley Creeks and were operated through September 2007. In October 1990, three rain gages were installed in each watershed and were operated through September 2007. Best-Management Practices (BMPs) were installed during 1993 to 2000 in Eagle and Joos Valley Creeks and were tracked throughout the study period. By the year 2000, a majority of the BMPs were implemented in the two watersheds and goals set by the Wisconsin Department of Natural Resources and the local Land Conservation Department had been achieved for the two study watersheds (Wisconsin Department of Natural Resources, 1990). The distributions of the rainstorms that produced surface runoff and storm loads were similar in the pre-BMP (1990-93) and post-BMP implementation (2000-07) periods for both Eagle and Joos Valley Creeks. The highest annual streamflow occurred at both sites in water year 1993, which corresponded to the greatest above normal nonfrozen precipitation measured at two nearby NOAA weather stations. The minimum streamflow occurred in water year 2007 at both sites. Base-flow and stormwater samples were collected and analyzed for suspended solids, total phosphorus, and ammonia nitrogen. For both Eagle and Joos Valley Creeks the median concentrations of suspended solids and total phosphorus in base flow were lower during the post-BMP period compared to the pre-BMP period and were statistically significant at the 0.05 significance level. The decrease in median concentrations of ammonia nitrogen at both sites was not statistically significant at the 0.05 significance level. Multiple linear regression analyses were used to remove the effects of climatologic conditions and seasonality from computed storm loads. For both Eagle and Joos Valley Creeks, the median storm loads for suspended solids, total phosphorus, and ammonia nitrogen were lower during the post-BMP period compared to the pre-BMP period and were statistically significant at the 0.05 significance level. The decreases in storm-load regression residuals from the pre- to the post-BMP periods for both Eagle and Joos Valley Creeks were statistically significant for all three constituents at the 0.05 significance level and indicated an apparent improvement in water-quality in the post-BMP period. Because the rainfall characteristics for individual storms in the pre- and post-BMP periods are likely to be different, separate pre- and post-BMP regressions were used to estimate the theoretical pre- and post-BMP storm loads to allow estimates of precent reductions between the pre- and post-BMP periods. The estimated percent reductions in storm loads for suspended solids, total phosphorus, and ammonia nitrogen were 89, 77, and 66 respectively for Eagle Creek and 84, 67, and 60 respectively for Joos Valley Creek. The apparent improvement in water quality is attributed to the implemented BMPs and to a reduction in the number of cattle in the watersheds.