A series of experiments examined ontogenetically the effects of context saliency and CS-US interval on the learning of CS-US and context-US associations. Although adult rats generally showed an inverse relationship between CS and context learning that depended upon CS-US interval and context saliency, the opposite relationship held for preweanlings, Consistent with many contemporary learning theories, adult rats expressed CS-context competition in learning about the US. Preweanlings, on the other hand, exhibited potentiated responding to either the context or the CS as a result of strong learning about the other element. These results suggest that adult and infant rats process information in a qualitatively different manner.Infants and adults may process information in a qualitatively different manner. For example, young animals do not show adults' inclination for overshadowing when they are tested on a single element of a compound training stimulus (Kraemer, Lariviere, & Spear, 1988;Mellon, Kraemer, & Spear, 1991) but rather show potentiation, the opposite of overshadowing (Hinderliter & Misanin, 1988;Kucharski & Spear, 1985;Spear & Kucharski, 1984). Age-related differences also exist in the treatment of contextual stimuli (Lariviere, Chen, & Spear, 1990; McKinzie, Lee, Bronfen, Spear, & Spear, 1994;Solheim, Hensler, & Spear, 1980). For example, McKinzie and Spear (1991) found that preweanling rats expressed strong learning to both context and tone following tone-footshock pairings. Context conditioning in animals given pairings ofa conditional stimulus (CS) and an unconditional stimulus (US) exceeded that in preweanlings given unpaired presentations of the same events. Thus, pairing of a tone CS with the US potentiated responding to the context (see also Goodwin et aI., 1992).Adults are more likely to attend to stimuli in the selective manner indicative of overshadowing. Punctate CSs usually are chosen for their capacity to elicit an orienting reflex (Pavlov, 1927) and, therefore, are likely to overshadow contextual stimuli in a typical conditioning experiment. Associative competition between the context and CS occurs in procedures that vary contingency and contiguity of the CS-US relationship. For example, Odling- Smee (1975) found that, as CS-US contingency decreased, conditioning to the training context increased. Marlin (1981) found that an inverse relationship existed between context and CS learning as a function of CS-US interval. When the CS and US were temporally contiguous, responding to the CS overshadowed responding to the context. As CS-US interval increased, however, conditioning to the CS diminished, while context learning increased. At an intermediate trace interval, conditioning occurred to both the context and the CS, whereas only context conditioning occurred at long CS-US intervals. Marlin concluded that both context and CS stimuli potentially form associations with the US and that learning occurs to the better predictor of the forthcoming US.The purpose of the present series of experiments...