This chapter describes some psychological theories that are at the foundation of research on cognitive models of emotions, and then reviews the most significant projects in this domain in recent years. The review is focused on probabilistic dynamic models, due to the key role of uncertainty in the relationships among the variables involved: the authors' experience in this domain is discussed by outlining open problems. Two aspects are discussed in particular: how probabilistic emotion models can be validated and how the problem of emotional-cognitive inconsistency can be dealt with in probabilistic terms.