We investigated the effects of transparent fraction names on children's reasoning about fractions. U.S. and Korean first and second graders were tested using verbal and nonverbal measures. On a verbal task, Korean students were worse at interpreting their own conventional fraction names than interpreting modified terms with a more familiar word for "parts." This indicates that Korean fraction names may not be as transparent as previously claimed (e.g., Miura et al., 1999). On nonverbal tasks, such as matching equivalent pictures of fractions and solving mixed-number problems using objects, students from both nations performed above chance and about equally well. However, U.S. students were more likely than Korean students to misinterpret a sharing task in terms of whole numbers, leading to significantly below-chance performance in the former group. (Contains 6 figures and 1 table.)