Tillage information is crucial in environmental modeling as it has a direct impact on soil erosion and water holding capacity of agricultural soils. A remote sensing approach is promising for the rapid collection of tillage information on individual fields over large areas. In this study, existing Thematic Mapper (TM)-based logistic regression models will be used to distinguish conventional and conservation tillage practices in Moore and Ochiltree counties located in the southern high plains of Texas. Accuracy assessments of tillage maps derived from Landsat TM data will be made using field data collected during 2005 planting season. The “percentage correct” and kappa (k) values will be used evaluate the accuracy of tillage maps.