Context <p>Advances in remote sensing and cloud-based computing have led to an increase in publicly accessible datasets characterizing the spatial extent of forested ecosystems. These datasets are used in ecological and environmental research, natural resource management, and policymaking. However, the number of available datasets and the often-subtle differences among them pose challenges for users seeking appropriate data for their specific objectives.</p> Objectives <p>We evaluate 27 data products derived from 12 publicly available datasets that quantify the distribution of “forests” in terms of tree cover and/or forest land use across the conterminous United States (CONUS), with temporal coverage ranging from 5 to 30&#xa0;years. We ask: How, why, and where do these datasets differ in their estimates of forest extent and change over time?</p> Methods <p>Using information aggregation, data visualization, and statistical tests, we compare and discuss area estimates and trends over time at the CONUS and state levels. To support dataset selection and interpretation, we developed an open-access map comparison tool.</p> Results <p>Estimates of the total area of forest ecosystems in CONUS differ by over 2,000,000&#xa0;km², and correlations among estimates vary in direction and statistical significance. State trend estimates are mixed and sensitive to differences in dataset definitions. Datasets with the same spatial resolution can vary in their suitability for a study area given other characteristics. Our results highlight the importance of dataset selection, understanding dataset characteristics, and visually comparing datasets in a study area prior to use.</p> Conclusions <p>Our findings underscore the need for careful selection and transparent reporting in analyses of cover, use, and other dimensions and attributes of forested ecosystems. Our comparison shopper’s guide and decision-support tool support users towards those ends.</p>