Several studies have shown that the NDVI calculated from NOAA-AVHRR data is related to annual rainfall and primary productivity in Sahelian areas. Such correlations, however, are affected by several environmental factors and have been tested only with data accumulated during rainy seasons, which is not ideal for the prediction of crop yield. In the present study a methodology of NOAA AVHRR data processing is presented which utilizes NDVI computed only in the first part of some rainy seasons and statistically takes into account the geographical variability in land resources and atmospheric conditions. From the first results of the application of the methodology in Niger, its potential has been shown both for environmental monitoring, and, more specifically, for crop yield assessment and forecasting.