This study examines a new method of assessing emotion understanding in preschoolers. Prior research has established that preschool is a critical time for emotion understanding development and that emotion understanding abilities are related to such important outcomes as social competence. Traditionally, measures of emotion understanding present multiple-choice questions that require children to select one emotion that is most likely to be elicited in various situations. However, this study proposes an alternative method in which children are asked to explain their answers on a subset of items. Their open-ended responses are then coded for quality of reasoning. Results establish preliminary evidence of the reliability and validity of the new assessment method. Notably, multiple regression analyses indicate that the coded emotion reasoning scores are a better predictor of social competence than scores obtained using traditional multiple-choice procedures. Limitations of the study and implications for future research are also discussed.