Introduction. Crediting of people has a firm position in last decade, but now there is a problem of the granted loans quality. So the study of the credit behavior, its mechanisms and structures is actual question. In article emphasis was placed on the methodological part of the problem, it was examined various statistical and econometric tools to study of the credit behavior. Methods. Using binary choice model determinants of credit behavior were identified, and a typical portrait of average borrower was made. Using cluster analysis the volume of consumer loans by region was investigated. By using time series analysis for the volume of granted loans SARIMA-model and forecast was estimated. Results. The most important determinants of credit behavior are age, sex, level of education, income, the loan experience in the past, the number of income sources. As a result of the cluster analysis Russian regions were divided into three homogenous groups that differ in terms of granted loans, the number of credit institutions, payable on the granted loans. Forecast of the volume of granted loans has shown that the volume of loans will continue to have a positive trend, despite the economic crisis.