The present experiment tested whether 7-month-old infants process a face as a configuration or as a collection of independent features. Thirty-two 7-month-old infants participated. Infants were habituated to two female faces and then tested on a familiar face, a switched face (consisting of all familiar features but a new combination of those features), and a novel face. It was reasoned that if infants processed only independent features of the face, the switched face should not look novel to them relative to the familiar face. If, however, infants processed the relationship among features (i.e. a configuration), the switched face should look novel. Furthermore, the study was based upon the well-established finding that inversion disrupts configural processing of faces. Thus, one group of infants saw only upright faces throughout the experiment, whereas a second group saw only inverted faces. It was hypothesized that by the age of 7 months, infants should process an upright face as a configuration, but an inverted face as independent features. The findings confirmed this hypothesis. In the upright condition, infants looked longer at the composite (switched) face than at the familiar face, whereas in the inverted condition infants did not. These findings are consistent with other reports of 7-month-old infants' ability to process configurations of both faces and non-faces and are consistent with the information processing approach proposed by Cohen (1998. An information-processing approach to infant perception and cognition. In The Development of Sensory, Motor, and Cognitive Capacities in Early Infancy, Simion F, Butterworth G (eds). Psychology Press: East Sussex; 277–300). Copyright © 2001 John Wiley & Sons, Ltd.