 This research proposes a new methodology for mapping annual crop lands at high spatial resolution without the need for field data. It uses existing baseline land cover information and temporal features specific to crops to produce accurate maps at different stages throughout the growing season. The methodology was tested at eight sites around the world and proved successful, achieving an overall accuracy of greater than 85%. Additionally, early maps could be generated at three-month intervals after the start of the growing season with accuracy increasing as the season progressed. This suggests that the methodology can be used to provide timely and accurate crop yield estimates which would be useful for operational agriculture monitoring programs. This article was authored by Nicholas Matten, Guadalupe Sepulcarcanto, Francois Waldner, and others.