The MACC / MACC-II projects (Monitoring Atmosphere Composition and Climate) have developed the new Heliosat-4 method to estimates the downwelling shortwave direct and global irradiances received at ground level under clear and cloudy skies. It is a fully physical model using a fast still accurate approximation of the libRadtran radiative transfer model serving as reference. It is composed of two abaci-based, or look-up tables, models: the already-published McClear model calculating the irradiance under clear sky and the new McCloud model calculating extinction of irradiance due to clouds. Inputs to Heliosat-4 are aerosol properties, total column content in water vapour and ozone produced by the MACC/MACC-II projects and cloud properties derived from images from the Meteosat satellites using the APOLLO (AVHRR Processing scheme Over cLouds, Land and Ocean) scheme. 15 min means of irradiance estimated by Heliosat-4 were compared to corresponding measurements made at 14 stations located in various climates within the Baseline Surface Radiation Network. The bias for global irradiance is comprised between 4 and 34 W m 2. The RMSE is around 90 W m 2 for most stations and its relative value ranges between 15% and 20% of the mean observed irradiance for stations in desert and Mediterranean climate, and between 26% and 45% for rainy climate with mild winters. The correlation coefficient ranges between 0.88 and 0.97. The bias for the direct irradiance is comprised between 44 and +46 W m 2. The RMSE ranges from 81 W m 2 (32% of the mean observed irradiance) to 189 W m 2 (64%). The correlation coefficient ranges between 0.81 and 0.95. Compared to other methods, this first version of Heliosat-4 reveals satisfactory results.