Two experiments explored repetition priming for familiar voices and faces. Expt 1 revealed that, like faces, prior exposure to a voice in a gender judgment task speeds its subsequent classification as familiar or unfamiliar, some minutes later. Faces and voices do not prime one another, however; a result consistent with the notion that evidence is initially accumulated separately for voices and faces. In Expt 2, a prediction derived from the IAC model of Burton, Bruce & Johnston (1990) was explored. The results confirmed that inter-modal repetition priming occurs when the interval between exposures to different personal identification stimuli are separated by a short SOA. This result is consistent with similar ones reported by Calder (1993) and Young, Flude, Hellawell & Ellis (1994) for face-name combinations.