Contextual salience modulates, but does not eliminate, the Other‑Race Effect.

The other‑race effect (ORE), often attributed to greater exposure to own‑race faces, reflects better recognition of those than other‑race faces. How experience would improve memory for new individuals remains unclear. We tested a cost‑efficient encoding hypothesis positing that when a salient feature such as skin colour makes a face distinctive, that feature receives disproportionate weighting during learning, reducing encoding of finer details. Manipulating distinctiveness based on ethnicity could thus potentially reverse the ORE. Using FaceGen, we created Caucasian and non‑Caucasian versions of 8 identities (4 targets, 4 foils) differing only in pigmentation. Caucasian participants (N = 96) learned four faces, one made distinctive by its skin colour (same‑race for half—SRD group; other‑race for the other half—ORD group). Pigmentation and distinctiveness status were counterbalanced. We tested recognition with skin colour removed to assess how much information beyond pigmentation was encoded and predicted reduced recognition for distinctive faces. Instead, same‑race distinctive faces were recognised better than the three other‑race faces in the SRD group, whereas no difference appeared in the ORD group. With skin colour restored, a distinctiveness advantage emerged in both groups. Therefore, while contextual saliency can modulate the ORE in a superficial way, representations for distinctive same‑race faces seem more robust, consistent with long‑term experience‑based tuning.