Across the globe, accurate national spatial datasets on cropland extent are lacking. These are necessary for a number of reasons, including accurately monitoring and predicting crop yield, land use, land acquisitions and food security. This study describes the use of crowdsourcing information retrieved over Ethiopia depicting the extent of cropland area. This information has been used to train a classification algorithm in Google Earth Engine to produce a continuous cropland extent map of Ethiopia. Preliminary results of this novel approach are encouraging, with an overall validity of 96%.