People are more accurate at recognizing faces of their own race than faces from other races, a phenomenon known as the other-race effect. Other-race effects have also been reported in some perceptual tasks. Across 3 experiments, White and Chinese participants completed recognition tests as well as the complete paradigm of the composite task, which measures participants’ abilities to selectively attend to the target region of a face while ignoring the task-irrelevant region of the face. Each task was completed with both own- and other-race faces. At a group level, participants showed significant own-race effects in recognition, but not in the composite task. At an individual difference level, the results provided no support for the hypothesis that a deficit in holistic processing for other-race faces drives the other-race effect in recognition. We therefore conclude that the other-race effect in recognition is not driven by the processes that underpin the composite effect.