Evaluation is a critical component of personal narrative, the component that conveys to listeners how narrators feel about experiences that happened to them. Evaluation conveys the impact of what actually did happen in the context of what narrators expected would happen but did not or what they wished had happened instead. This paper presents a study of how Taiwanese children develop the ability to evaluate their narratives and a comparison of Taiwanese to English-speaking children in their use of evaluative devices. Prior research (Minami, 1994; Minami & McCabe, 1995) suggests that English-speaking mothers provide more evaluation comments in telling past experiences with their children compared to Japanese-speaking mothers. Differences in use of evaluative devices were hypothesized to be evident across Chinese- and English-speaking children in their personally experienced stories. Mandarin Chinese-speaking children from Taiwan (N = 171) and 96 English-speaking children from the United States participated in this study. Chinese-speaking children were agesd 3 to 9 years and comprised seven groups (at each of those ages). English-speaking, American children were aged 4 to 9 years and comprised six groups. Following Peterson and McCabe’s conversational map (1983), the experimenter elicited a number of personal narratives from each child. Specific prompted topics such as visit to a doctor were given in conversation. The purpose of this task was to assess children’s narrative skill without adult’s support; neutral follow-up responses such as “uh-huh,” “tell me more” were used. Evaluation was coded in Chinese using an adaptation of the system Peterson & McCabe (1983) developed to code evaluation in English-speaking American children aged four through nine years. The percentage of each type of evaluative device per narrative comment was determined and Taiwanese children were compared in this way to American children from the Peterson & McCabe corpus. Results show that Taiwanese children included many fewer evaluation comments (13–25% of the children’s clauses were partially or fully evaluative) in telling their personally experienced stories compared to American children (50% at each age). Results are interpreted as reflecting deep and pervasive cultural differences: Chinese children are socialized to have an interdependent self (with less emphasis on what an individual felt in the past), while American children are socialized to have an independent self (with early and frequent emphasis on what an individual felt in the past).