We present evidence that the bald eagle (Haliaeetus leucocephalus) productivity survey in the Boundary Waters Canoe Area Wilderness of northeastern Minnesota may have underestimated the number of nestlings during 1986-1988. Recommendations are provided to achieve more accurate ground-based counts. By conducting ground-based observations for up to 1 hour/nest, an accurate count of the number of bald eagle nestlings can be obtained. If nests are only observed for up to 30 minutes/nest, an accurate determination of nest success can be made. The effort that managers put into counts should be based on the intended use of the productivity data. If small changes in mean productivity would trigger management action, the less acurate ground-based counts should be conducted with caution. Prior to implementing ground-based counts, a study like ours should estimate bias associated with different survey procedures and the observation time needed to achieve accurate results.