BU Kogalymneftegaz offers strong factual evidence allowing developing a methodology to predict oil and gas po- tential BY constructing probabilistic and statistical models based on geological zones description. A special feature of this method is that only variables permanently available to producers are considered, in addition, they will be used in- tegrally, providing high reliability of obtained probabilistic and statistical prediction models of oil and gas potential.This method of oil and gas potential prediction can be performed with certain zone maturity, i.e. when a number of zones is used for analyses, some of which contain hydrocarbons in the sediment and some are hydrocarbon-free. All of these zones can be characterized by the same parameters, and these parameters can be determined before apply- ing them for further research. This helps to build probabilistic and statistical models that can be practically used in as- sessing oil and gas potential of the undrilled areas. Such method allows planning the sequence of research within the areas. It is suggested that identification of the most promising oil and gas bearing zones should be made by construct- ing geological and mathematical prediction models. It is important that in constructing the models the probabilities cal- culated on the basis of parameters, but not the different dimensional parameters themselves will be used. To obtain the necessary probabilities regression equation will be constructed. Probability values used in incremental linear dis- criminant and multivariate regression analysis will permit to develop a complex probabilistic criterion. This criterion will be used in future to build a multidimensional model directly for the zones, on the basis of which it is possible to identify priority areas recommended for further research.