A review is given of the current state of the art in the prediction of rain attenuation on slant paths. Different types of research effort are identified and discussed. A critical assessment of data analysis methods and theoretical models is made to identify areas where more effort is needed. Various types of data available are discussed (direct attenuation measurements, radiometric data, rain rate, and radar reflectivity) as well as the models and methods available to date to relate these to attenuation on a satellite path. Finally, prediction methods for worldwide use are reviewed. While great progress is being made in the physical modeling of the atmosphere, statistical prediction methods are still very general and coarse. In the future, the emphasis should be on the development of detailed models to predict the variability of attenuation statistics within a coverage area. In this connection there is a need for coordination in the comparison of experimental data. In Europe, projects of international cooperation for this purpose are being carried out.