Researches support by competitive funding mechanisms realized through scientific foundations is a common global practice. At the same time, decisions on support or refusal of a scientific project financing are made because of results of a multistage examination, which is a primary part of the competitive funding system and is conducted by the scientific community itself (peer review). It is important to take into account that the decisions of each expert in a situation of choice are influenced by his individual economic and psychological characteristics. Usually, these characteristics are neglected, but in our opinion, they should not be underestimated. This article presents an agentbased model of professional expertise and decision-making on financial support for research projects by scientific foundations, considering the economic and psychological characteristics of agents-experts and the reputational consequences of their decisions. The model takes into account such individual economic and psychological characteristics of scientists as “individualism – collectivism”, “satisfaction – dissatisfaction” and others. The quality of projects, the qualifications of scientists, their propensity for subjectivity and objectivity in the process of project evaluation, depending on the economic and psychological characteristics are also considered. Feedbacks in the model are implemented by changing the reputation of scientists, depending on the degree of objectivity of projects evaluations including by comparing the reputation of the researcher with the average reputation of his immediate surrounding. Conclusions about changes in the structure of researcher’s groups within the scientific community are made on the basis of changes in the reputation of scientists, depending on their belonging to the class of dependent (pursuing the interests of individual groups and communities) or the class of independent (conducting a fair evaluation of research projects). Recommendations for the further development of the model and for using it to predict the outcomes of local situations in autonomous socio-economic systems are formulated.