This study examined whether inducing a general relational mindset could reduce retrieval-induced forgetting (RIF). In three experiments, a relational mindset was induced by solving cross-domain analogy problems, with a within-domain analogy task serving as the control, followed by a retrieval test for all participants. The results revealed that, regardless of whether the task was an item recognition test (Experiment 1) or a classical retrieval practice paradigm (Experiment 2), participants in the relational mindset induction group showed no significant difference in recalling nonpracticed items (Rp−) with baseline items (Nrp), whereas participants in the within-domain analogy task control group exhibited significant retrieval-induced forgetting. Experiment 3 extended Experiment 2 by adding a no-task control group that directly underwent a retrieval-practice task. The results confirmed that retrieval-induced forgetting was observed in both the within-domain analogy task control and no-task control groups but was eliminated in the relational mindset induction group. Additionally, supplementary trial-level analyses in Experiments 2 and 3 indicated that higher performance on the induction task was associated with better recall of Rp− items, suggesting that a strong relational mindset may even lead to retrieval-induced facilitation. These findings suggest that inducing a general relational mindset could reduce retrieval-induced forgetting. In addition, the general relational mindset is unrelated to the specific content of learning, allowing educators to apply this approach in some cases to effectively reduce RIF.