Initiated as part of GEOGLAM (Group on Earth Observations Global Agricultural Monitoring), the ECYFS (Earth observation-based Crop Yield Forecasting System) project aims at designing an innovative yield-forecasting methodology, based on an algorithm of machine learning combined with satellite and meteorological data, and a culture model. ECYFS will be developed for two major agricultural commodities: wheat and corn. It will be calibrated and validated on a wide range of agricultural regions (mainly European Union countries) characterizing cropping systems and contrasting weather and environmental conditions.