In the past decades, affective science has overwhelmingly demonstrated the unique properties of affective information to bias our attention, memory, and decisions. At the same time, accumulating evidence suggests that neutral and affective representations rely on the same working memory substrates for the selection and computation of information and that they are therefore restricted by the same capacity limitations that these substrates impose. Here, we integrate these insights into a working memory model of affective processing (WMAP). Drawing on competitive access (...) models of working memory, we discuss its role in the various stages of affective processing, from attentional selection to maintenance and memory storage, and resulting feelings and actions. We end our overview with some open questions and future directions. (shrink)