The paper discusses the development and characteristics of the National Land Cover Data Set (NLCD) for the conterminous United States, derived from Landsat Thematic Mapper (TM) data and ancillary sources. The NLCD, completed in 2000, represents an intermediate-scale national land cover dataset with a 30-meter resolution, suitable for regional to national-scale investigations. The dataset includes 21 land cover classes and is designed to be consistent and seamless across the country. The methodology involves terrain-corrected TM data, ancillary spatial data, and a hierarchical classification system. Accuracy assessments were conducted using aerial photographs, achieving an overall accuracy of 81% for Anderson Level I classes and 60% for all classes. The NLCD is available for download and has been widely used by federal agencies, state agencies, universities, and non-governmental organizations for various applications. However, the dataset is not recommended for local-scale analyses due to its coarse resolution and potential classification errors. Future work aims to improve the dataset's accuracy and flexibility, potentially using Landsat 7 data and more detailed ancillary information.The paper discusses the development and characteristics of the National Land Cover Data Set (NLCD) for the conterminous United States, derived from Landsat Thematic Mapper (TM) data and ancillary sources. The NLCD, completed in 2000, represents an intermediate-scale national land cover dataset with a 30-meter resolution, suitable for regional to national-scale investigations. The dataset includes 21 land cover classes and is designed to be consistent and seamless across the country. The methodology involves terrain-corrected TM data, ancillary spatial data, and a hierarchical classification system. Accuracy assessments were conducted using aerial photographs, achieving an overall accuracy of 81% for Anderson Level I classes and 60% for all classes. The NLCD is available for download and has been widely used by federal agencies, state agencies, universities, and non-governmental organizations for various applications. However, the dataset is not recommended for local-scale analyses due to its coarse resolution and potential classification errors. Future work aims to improve the dataset's accuracy and flexibility, potentially using Landsat 7 data and more detailed ancillary information.