In this paper the crop water stress is evaluated with three different indices: (i) the Relative Soil Moisture Index [RSMI] resulting from agrometeorological model simulations, (ii) the Normalized Difference Vegetation Index, [NDVI] applied on S10 SPOT-VGT imagery, (iii) the Normalized Difference Water Index, [NDWI] applied on S10 SPOT-VGT imagery. The RSMI is one of the outputs of the Belgian Crop Growth Monitoring System. It indicates the soil water availability for crops. Crop water stress is assumed when soil water availability is lower than the crop potential evapotranspiration. One of the potential applications for drought monitoring is the calculation of the number of days with a crop water stress during the growing season or more precisely during the “moisture-sensitive period” of the crop. The two others (well-known) indices are derived from low resolution satellite imagery. Complete time series of S10 SPOT-VGT NDVI and NDWI data - i.e. from April 1998 onwards – have been acquired for the research. Using spatial information on land-uses in Belgium, a simple unmixing method is implemented to partly reduce the problem of mixed values occurring inevitably in 1x1km² pixels. Only the most agricultural pixels are considered for the analysis. The heat wave occurred during the summer 2003 has been chosen as a study case for the comparison. This summer was particularly hot, dry and sunny with the highest averaged temperature ever recorded from June to August in Uccle-Belgium. Many human activities were affected by this particular climatic condition, among which agriculture. In order to make the comparison possible, a standardization of the outputs of the 3 indexes is realised, taking into account the times series of each indexes. Moreover all the outputs are aggregated at municipality level.