Reflective equilibrium (RE) is often taken to be a powerful method in ethics, logic, theories of rationality, or philosophy in general. But so far, RE has never been characterized in a rigorous way. This talk presents a formal model of the RE. I focus on the key idea of RE, i.e. the elaboration of an agent's commitments under the pressure from systematic theories. This idea is spelled out in the formal framework of the Theory of Dialectical Structures (Betz, 2010, 2012). Commitments and theories are represented using dialectical structures; desiderata on sets of commitments and theories are introduced; a state of RE is defined, and rules for adjusting commitments and theories are specied. This model then is tested in applications to simple examples, which lend intuitive support to our model. I conclude by discussing the assumptions underlying our model as well as its limitations.