The Cross-Race Effect (CRE) is a robust finding in social cognition. One reason the CRE arises is due to the relative expertise people have with same-race compared to cross-race faces. Here, we examine the effects of reducing such expertise through blurring same-race and cross-race faces. White participants exhibited the CRE for clear faces, but this pattern was reversed for blurred faces (Study 1). Participants remembered Black faces as having been presented more clearly than White faces (Study 2), which held specifically for blurry faces (Study 3), suggesting that a clearer representation of blurred Black faces created superior recognition memory, an assumption confirmed in Study 4. This clarity bias also existed in short-term memory (Study 5) and was specific to faces (Study 6). This research advances theoretical accounts of the CRE and highlights the discrepancy between how the effect is often studied and how it may function in real-world settings.