An interactional model of stress that integrates current research on competitive affects and emphasizes the temporal dimensions of the stress process is forwarded. The literature reveals that the study of athletes' affective responses to competition has been narrowly focused on pre-competitive anxiety. Equivocal findings on temporal patterning of competitive anxiety suggest that a fundamental change in the empirical approach is needed because the current conceptualization of anxiety and other complex emotions is imprecise. The analysis of secondary emotions as patterns of discrete basic emotions, as suggested by differential emotions theorists, is proposed for consideration in future research. In this view, competitive anxiety is considered as a set of patterns of emotions rather than a unitary affect. The adoption of this approach could result in better operationalization of competitive anxiety as well as other secondary performance-related emotions. We propose that research on competitive affects should follow two parallel lines. The first should focus on the description of complex emotional states that reflect the idiosyncratic emotional experience and vocabulary of the athlete. The second should examine the sets of basic emotions experienced throughout competition, and focus on individual differences and factors determining those differences. The integration of the two approaches could lead to a better understanding of whether, how and why individuals differ in the interpretation of specific secondary emotions and their effect on performance. Moreover, it would permit the analysis of intra-individual variations in labelling secondary emotions with respect to different competitive contexts and temporal aspects.