Investigators of cognitive development have historically relied nearly exclusively on studying how the infant behaves. This has limited our knowledge of this area in several important respects. For example, the link between cognition and behavior is not always clear, both because the infant is a motorically immature organism and because what the infant knows may not always be reflected in how the infant behaves. More importantly, we know from behavioral studies that infants attain certain milestones in their cognitive growth over the first year of life (e.g., object permanence). Although it is often assumed that these behavioral changes are brought about by corresponding changes in the brain, behavioral measures alone are unable to elucidate this relationship. For these and other reasons it seems imperative to move beyond the study of behavior to include, as well, the study of underlying physiological activity. Because of the critical role the brain plays in mediating cognitive processes, it seems reasonable to propose that the brain be the subject of this study. It is to this end that the current proposal is directed. Specifically, long-latency Event-Related Potentials (ERPs) will be recorded from the scalp in 4- to 12-month-old infants as a means of studying infant memory. In order to relate changes in the brain to changes in infants' behavior in several of the proposed studies behavioral measures will be recorded as well. The first two experiments are designed to extend previous research by the P. I. (Nelson & Salapatek, 1986), by examining infants' cortical and behavioral responses to stimulus noveity (Experiment 1), and infants' cortical responses to a stimulus omission (Experiment 2). The third experiment examines the relationship between infants' ERP response to novelty, and subsequent recognition memory as assessed behaviorally. Experiment 4 has as its focus infants' memory for changes in stimulus duration, and in the final experiment the electrophysiological correlates of cross-modal transfer are examined. Collectively the goal of this proposal is to study infant memory through the examination of both cortical and behavioral response systems.