Previous research has shown that category expectations can improve recall, by reducing absolute average error (e.g. Hut-tenlocher, et.al., 1991; Hemmer & Steyvers, 2009), particularly when expectations are consistent with studied information.However, studied information that is expectation-inconsistent may also boost memory (e.g. Sakamoto & Love, 2004).Using a cued-recall task, we manipulated the degree to which studied object-color pairs aligned with peoples (N=29)expectations to explore the role of expectations in delayed recall. Our preliminary results show greater recall accuracyfor expectation-consistent items (e.g. yellow bananas) compared to expectation-inconsistent (purple bananas), and no-expectation items (yellow toothbrushes). However, there was no difference in accuracy between expectation-inconsistentand no-expectation items, nor was there a difference between weak and strong expectation-inconsistent items (orangish-yellow and purple bananas, respectively). This preliminary work shows that in delayed recall, the benefit of categoryexpectations might not extend to instances when studied information is misaligned with those expectations.