▎ 摘　　要

Background: The prediction of physicochemical properties is important task of the natural sciences. Quantitative structure - property relationships (QSPR) are a tool to solve the task. Objective: QSPR for dispersibility of graphene in various organic solvents has been built up by means of the CORAL software (http://www.insilico.eu/coral). Method: The Monte Carlo technique is the basis of the models for dispersibility of graphene in various organic solvents. Simplified molecular input-line entry systems (SMILES) are used to represent the molecular structure for the QSPR analysis. In other words, the graphene dispersibility is modeled as a mathematical function of the molecular structure. Results: The statistical characteristics of the models are quite good. They have the mechanistic interpretation: the structural features of molecules of solvents which are promoters of increase or decrease of graphene dispersibility have been discovered. Conclusion: The suggested approach can be used to predict dispersibility of graphene in various organic solvents.