The article is devoted to the analysis of river runoff data provided by the latest generation of the Coupled Atmosphere-Ocean General Circulation Models (AOGCMs) in the framework of the CMIP5 project. The main aim of this research is to forecast the probable changes in river water resources of the main rivers of the Russian Federation. At the first stage of investigation, a comprehensive analysis of the river runoff computational models and the quality of the initial AOGCMs information were carried out. This analysis was performed in several steps. First, the average long-term model river runoff for the base period of 1981-2000 was compared with the average long-term values of the observed discharge for the seven largest rivers of the Russian Federation (Volga, Pechora, Northern Dvina, Ob, Lena, Yenisei and Amur). In addition, the spatial distribution of the river runoff throughout the territory of Russia provided by the AOGCMs was analyzed. Then, the adequacy (negative or unrealistic runoff values) of the si