The strength of most Russian psychological research has been and remains the presence of a preliminary theoretical foundation for the hypotheses put forward. The use of this methodological approach in the study of a person's subjective well-being, especially conducted with the use of advanced mathematical and statistical methods, as well as algorithms of linear and nonlinear machine modeling, requires the presence of an appropriate pre-constructed conceptual explanatory model. One area of research on subjective well-being in which such models are absent, or at least not widely present, is the study of subjective well-being predictors. The aim of this article is to develop a theoretical research model that will allow us to distinguish between the personal, process, and contextual levels of predictors of subjective well-being, and to separate the appearance of this phenomenon from other phenomena. A brief theoretical analysis of the mechanisms of functioning and grounds for the sustaining of human subjective well-being, based on the concept of subjective well-being as a complex dynamic system, is presented. The results of this analysis are used to develop a theoretical research model for studying predictors of subjective well-being. In the suggested model, the components of subjective well-being are considered as a set of human reactions to a person's activity, and these reactions themselves as a product of dynamic nonlinear interaction of personality aspects, mental processes and socio-psychological context of human activity. The proposed research model can be used both for preliminary theoretical grounding and further empirical testing of hypotheses put forward in the study of predictors of subjective well-being, and for comparison of the results obtained with the findings of other studies.