||Evidence that the Conservation Reserve Program (CRP) has resulted in large-scale increases in populations of grassland birds is limited. Detecting large-scale CRP effects is difficult because agricultural landscapes are complex, dynamic systems where many concurrent changes are occurring across space and time, and CRP is only one of many factors influencing wildlife populations. Trying to isolate and quantify the contribution of CRP to large-scale population changes under these conditions is extremely difficult and tenuous. Data-quality issues affecting many large-scale monitoring programs exacerbate the problem. We use a case study of land-use and pheasant-monitoring data in Minnesota from 1974-1997 to illustrate these problems. In our example, roadside counts of Ring-necked Pheasants (Phasianus colchicus) were correlated positively with percent of CRP grasslands within 1.6 km of survey routes, but the predicted change in mean pheasant counts (pre-CRP vs. CRP) was negative in three of five regions despite the addition of up to 8% CRP grasslands. We also documented concurrent losses (1.8%-6.1% per year) of alternative reproductive habitats that apparently counteracted the positive association between pheasant counts and CRP abundance. These results illustrate the need for a more comprehensive evaluation of Farm Bill effects on wildlife, including commodity provisions that lead to conversion of pasture, hayland, and small grains to row crops.