The paper presents modern data on the distribution of Ephedra distachya L. an attempt was made to model the bioclimatic range of the species using the maximum entropy method and compare the obtained data with the range defi ned in the report «Flora of the USSR», taking into account additions from the monographs «Flora of the European Part of the USSR» and «Geography of Trees plants of the USSR». MaxEnt modeling of the potential range of E. distachya was carried out on the basis of the herbarium collections of Saratov State University (SARAT), the GBIF database and the Plantarium electronic atlas. A total of 6973 species location points were used. In accordance with the obtained bioclimatic model, in the modern climate the most favorable conditions for the existence of the species are on the coasts of the Black, Azov, Marmara, Mediterranean and Caspian Seas, in the territories of the former AzSSR, Ukrainian SSR and Georgian SSR. In the RSFSR (modern Russian Federation), the species is optimally located in the Volga, North Caucasus and Southern federal districts. A sharp narrowing of the potential range is observed in the Ural and Central Federal Districts of the Russian Federation, as well as in the territories of the former TurSSR and KazSSR. The accuracy of the model is confi rmed by the high AUC (Area Under Curve) score, which is 0.933 for training data and 0.930 for test data. Climatic parameters have been established that infl uence the distribution of the species, in which its existence is optimal. The greatest contribution to the construction of the model after permutation was made by isothermality (32%), average annual temperature (23.1%), average monthly daily temperature amplitude (12.4%), as well as the maximum temperature of the warmest month (11.6%). A correlation analysis of the parameters that made the greatest contribution after permutation was carried out. When estimating the error using the jackknife method, a variable (average annual temperature) was obtained that contains the most information that is not in other variables; the exclusion of this variable leads to deterioration of the model.