Analogical reasoning, the mapping of structured relations across conceptual domains, is commonly recognized as essential to human cognition, but young children often perform poorly in the classical A:B::C:? analogical reasoning task. Particularly, young children have trouble when the objects in the task are not strongly associated with each other, and/or when there are strong associative lures among the potential answers. Here, we examine whether successive trials that repeat the same relation needed to solve the analogy can help overcome some of the challenges with weakly associated items. In the first of two experiments, our results were mixed. In the second, we simplified the design, and were able to more clearly show a benefit of repeating relations across consecutively solved problems.