The current study tested the widespread assumption that holistic processing is important for the identification of upright faces. In two experiments, we show that (a) there are large individual differences in the magnitude of the composite face effect and face-identification accuracy and (b) the correlation between the magnitude of the composite face effect and face-identification accuracy is essentially zero. These findings are inconsistent with the claim that holistic processing, as indexed by the composite face effect, significantly influences accuracy in a face-identification task.