Recent studies have used self-report methods to defend a close associative or causal connection between appraisal and emotion. The present experiments used similar procedures to investigate remembered experiences of reasonable and unreasonable anger and guilt, and of nonemotional other-blame and selfblame. Results suggest that the patterns of appraisal reported for reasonable examples of emotions and for situations where there is a near absence of emotion may be highly similar, but that both may differ significantly from the appraisal profiles reported for (...) unreasonable examples of the same emotions. Further, relevant appraisals were not always identified by participants as the most influential determinants of guilt and anger. These findings demonstrate either that the relationship between certain appraisals and emotions is less consistent than implied in some contemporary versions of appraisal theory, or that there are problems with the validity of existing questionnaire-based measures of the variables in question. (shrink)