We tested 4–6‐ and 10–12‐month‐old infants to investigate whether the often‐reported decline in infant sensitivity to other‐race faces may reflect responsiveness to static or dynamic/silent faces rather than a general process of perceptual narrowing. Across three experiments, we tested discrimination of either dynamic own‐race or other‐race faces which were either accompanied by a speech syllable, no sound, or a non‐speech sound. Results indicated that 4–6‐ and 10–12‐month‐old infants discriminated own‐race as well as other‐race faces accompanied by a speech syllable, that only the 10–12‐month‐olds discriminated silent own‐race faces, and that 4–6‐month‐old infants discriminated own‐race and other‐race faces accompanied by a non‐speech sound but that 10–12‐month‐old infants only discriminated own‐race faces accompanied by a non‐speech sound. Overall, the results suggest that the ORE reported to date reflects infant responsiveness to static or dynamic/silent faces rather than a general process of perceptual narrowing. Prior studies have found that discrimination of other‐race faces declines in infancy. This other‐race effect (ORE) has been found in studies of infant discrimination of static or silent dynamic faces. We posited that the ORE found to date may reflect the specific nature of the stimuli used previously rather than infants’ typical social experiences with talking faces. To examine this prediction, we tested 4‐6 month‐old and 10‐12 month‐old infants’ discrimination of own‐race and other‐race faces that could be seen and heard producing a speech sound. We found successful discrimination of both types of faces at both ages, indicating that infants maintain their sensitivity to other‐race faces into the second year of life as long as the faces are dynamic and specified by redundant multisensory perceptual cues.