The goal of this experiment was to investigate elementary school children’s ability to acquire basic fraction knowledge. The degree of concreteness of instantiations of proportions was varied between subjects. First-grade children learned to label proportions of objects with fraction. Proportions were presented either as concrete, colorful flowers or as generic black and white circles. Following instruction, participants were given a test of learning and an immediate or delayed test of transfer involving proportions of novel objects. Those who learned with the generic materials scored higher on learning and transfer than those who learned with the concrete materials. Differences between learning conditions were attenuated for the delayed transfer test. These findings suggest that concrete, perceptually rich instantiations of fractions may hinder children’s acquisition of basic fraction knowledge in comparison to simple generic instantiations of fractions.