The current research starts from the idea that variance is at the heart of emotion. Rather than assuming that individuals always experience an emotion like anger or shame in the same way, our starting point was that anger or shame experiences may vary from one instance to another. We expected to find some regularity among these heterogeneous instances, such that there are different anger and shame experience types, that is, groups of people who experience similar instances of anger and shame. Studying cultural differences in emotional experience then means studying differences in the distribution of these types across cultural contexts. Students from the United States, Japan, and Belgium (N = 928) indicated their emotional experiences in terms of appraisals and action tendencies in response to 15 hypothetical anger or shame situations. Using an inductive clustering approach, we identified anger and shame types who were each characterized by a specific pattern of anger and shame experience. As expected, we found that the distribution of these types differed across the three cultural contexts: Of the two anger types, one was common in Japan and one in the U.S. / Belgium; of the three shame types, one was most prevalent in each cultural context. Participants’ anger and shame type was primarily predicted by their culture of origin, and not, or much less, by their ethnicity, SES, gender, self-construal, or personality. We discuss the value of this approach in overcoming (some of) the essentialist constraints that have constrained cultural psychological research on emotion.