The point of the article is to show possibilities and limits of individual and group social behavior sociological modeling method. In the article model of social behavior as a simplified representation of reality is considered, some kind of quasi equivalent of reality. In the article a new approach to the evaluation of satisfiability of the behavioral model according its aim and reasons are presented. Underlying reasons (particular tasks) of behavioral models consists in definition, prediction, interpreting and regulation. To solve the above mentioned problem of definition one can make a model of social processes using method of unfinished phrases. To solve the problem of prediction modeling of social processes on the base of experiment is suggested. The problem of regulation by constructing of normative model of behavior on the base of repertory grid technique (J. A. Kelly) can be solved. Practical cases for modeling and using of above mentioned methods for different research situations are listed. Possibilities and limits of each of the approach are viewed. Two models of expected behavior are shown: at situation of certainty and risk situation. Ways of empirical data interpreting logic in terms of structural and motivational approaches are defined. Author concludes that results of the correct appointing of theoretical and methods aspects of sociological modeling of behavior one can get information about content, variability and differences in social behavior of individuals and groups.