Three shaved adult male albino rats were trained to press a lever to replace a hot air drive condition (50°C) with either a small reinforcer (32°C) or a large reinforcer (14°C). Following the 10-sec reinforcer, the drive condition was either reinstated immediately (no delay) or after a 15-sec exposure to the drive temperature, during which the bar was withdrawn (delay). Response rate during the no-delay procedure was faster for the small reinforcer than for the large. This relation reversed during the delay procedure. The former observation is similar to a satiation effect and the latter resembles an incentive effect.The effects of thermal stimuli on instrumental behavior have been investigated primarily by students of physiological regulatory systems. Consequently, many traditional questions about the behavioral properties of motivational events have not been asked about thermal stimuli. The present work determines the role of "incentive" in thermally motivated behavior. Incentive is usually identified as an increase in response strength associated with an increase in reinforcement magnitude (Crespi, 1942;Zeaman, 1949;Guttman, 1953).In the thermal equivalent of the reinforcement procedure, a rat is trained to press a bar to terminate a stressful "drive" condition and replace it with a stress-reducing "reinforcer." The reinforcer is terminated automatically and replaced by the drive condition for the next trial. The drive and reinforcer are defined here as stimulus values and not as intervening variables such as thermal stress. Although it might seem that thennal stress could be quantitatively defined as the internal temperature of the animal, it is lamentably true that internal temperature is actually a temperature gradient. A temperature drop of 5°C occurs between the core, which is relatively invariant, and the skin, which is strongly affected by both air temperature and vasotone. The formulation of an internal index of stress requires a greater understanding of the physiology of heat transport than is currently available. For present purposes, it will suffice to specify treatments in terms of air temperature and assume that thermal stress is some function of air temperature.Th e motivational parameters of the thermal reinforcement procedure have been studied with unusually reliable results. As the intensity of either a hot or cold drive stimulus is increased, response rate