Deprivation shifts, using conditions analogous to those which produce strong incentive contrast when reward quantity is changed (extended preshift training, short interval from the last preshift trial to the first postshift trial, large preshift differences in deprivation) resulted in contrast of runway speed and choice behavior, In the first experiment, a downshift of hunger during rats' runway training produced a slow lessening of speed below that of a group trained continually at low hunger. In the second experiment, rats were trained to traverse a runway to one food-containing goalbox when very hungry and to another, distinctively different food-containing goalbox when not very hungry. The rats were next given a series of choice trials between the two goalboxes. There was a brief preference for the high-hunger goalbox, followed by a preference for the low-hunger goalbox. The results of the second experiment suggest that deprivation affects the strength of conditioning of a cue-reinforcer expectancy, while the slow development of contrast in the first experiment indicates that deprivation also affects the development of either habit strength or a response-reinforcer expectancy.Rats switched from high to low hunger following a small number of rewarded runway trials run faster on initial rewarded postshift trials than do animals trained all along in the state of less hunger (Brush, Goodrich, Teghtsoonian, & Eisman, 1963 Mollenauer, 1971). Associative and incentive-motivational interpretations of this result have been offered. One associative explanation assumes that the level of deprivation at which a stimulus-response sequence is followed by consumatory stimulation affects the asymptotic value of the stimulus' tendency to evoke the response (Capaldi, 1972; Eisenberger et aI., 1973;Hovancik, 1978). Incentive-motivational interpretations hold that the level of deprivation at which consumatory stimulation occurs affects the strength of conditioning of reward expectancy to the exteroceptive and proprioceptive stimuli that precede the reward (Black, 1965; Bolles, 1958Bolles, , 1975 magnitude is decreased than when the lesser reward magnitude is employed all along. As with the downshift of deprivation level, the initial performance of animals receiving a reduction in reward during runway training is better than that of a group consistently receiving small reward. However, speed soon falls to a level less than that of the constant-smallreward group. Incentive motivational interpretations attribute this negative incentive-contrast effect to the occurrence of a reinforcing event that is of lesser magnitude than anticipated (Black, 1965;Bower, 1961). Failures to find contrast following downshifts of deprivation fail to support a simple analogy between shifts of reward quantity and shifts of the level of deprivation during instrumental performance.Mollenauer (1971) pointed out that most deprivationshift experiments included relatively few preshift trials, although the magnitude of contrast produced by the downshift of r...