To adequately monitor Neotropical migratory birds, information must be collected to assess population change at local, regional, and continent-wide scales. I suggest that large-scale survey results (such as those derived from the North American Breeding Bird Survey) should not be used to predict population attributes on parks, refuges, and other protected areas. These areas are often managed, and generally contain habitats that can be poorly sampled in large scale surveys, hence local bird populations might be quite different from those sampled in the large-scale surveys. Furthermore, we are limited in our capabilities to combine information from local surveys with large-scale survey data. Most surveys of bird populations collect indices of abundance which are often not comparable among surveys due to habitat and region specific differences in probabilities of detecting birds. In assessing the effects of management, it is important to understand the limitations of monitoring at different geographic scales and to design programs to monitor at the scale at which management is conducted.