A fundamental question in affective sciences is how the human mind decides if, and in what intensity, to elicit an affective response. Appraisal theories assume that preceding the affective response, there is an evaluation stage in which dimensions of an event are being appraised. Common to most appraisal theories is the assumption that the evaluation phase involves the assessment of the stimulus’ relevance to the perceiver’s well-being. In this talk, I first discuss conceptual and methodological challenges in investigating relevance appraisal. Next, I present two lines of experiments that ask how the human mind uses information about objective and subjective probabilities in the decision about the intensity of the emotional response and how these are affected by the valence of the event. The potential contribution of the results to appraisal theory is discussed.