Pigeons were trained on either a variable-interval 60-second schedule, or on a schedule that differentially reinforced responses that were spaced at least 20 seconds apart. The birds were then exposed to several durations of reinforcement delay, with comparisons between signaled and unsignaled delays. Although unsignaled delays of 5 and 10 seconds produced large decreases in response rate, signaled delays of up to 10 seconds produced only moderate decreases in response rates. In addition, some subjects responded more rapidly with a .5 or 1.0 second duration of unsignaled delay than with immediate reinforcement. These response rate changes occurred regardless of whether the rate of reinforcement concomitantly decreased or increased.Key words: reinforcement delay, variable-interval schedule of reinforcement, differentialreinforcement-of-low-rates schedule of reinforcement, key peck, pigeonsOne of the most widely discussed parameters of reinforcement is delay of reinforcement (e.g., Bolles, 1975;Hulse, Deese, & Egeth, 1975;Mackintosh, 1974). While much of the early empirical and theoretical work examined the manner in which delayed reinforcement influenced the learning of new behavior or discriminations (e.g., Grice, 1948;Perin, 1943;Spence. 1947;Wolfe, 1934), the more recent emphasis, especially in the area of operant conditioning, has been on how it affects behavior maintained by various schedules of reinforcement (e.g., Azzi, Fix, Keller, & Rocha e Silva, 1964;Dews, 1960;Ferster, 1953;Ferster & Hammer, 1965;Morgan, 1972;Pierce, Hanford, & Zimmerman, 1972;Silver & Pierce, 1969;Sizemore & Lattal, 1977Williams, 1976).Delayed reinforcement actually refers to a collection of very different manipulations. For example, the delay period may be signaled or unsignaled; if signaled, the opportunity to respond may or may not be withdrawn during the delay. In addition, responding during the delay interval may be nonfunctional or it may be penalized, e.g., by resetting the delay interval. Given this diversity, it would be surprising if all such manipulations had identical effects. Yet, few researchers have actually compared different types of delayed reinforcement suggestion, Richards and Hittesdorf (1978) found that a 10-sec signaled delay maintained much higher response rates than a 10-sec unsignaled delay when pigeons' responding was reinforced according to a VI schedule.The purpose of the present experiment was to assess the generality of Richards and Hittesdorf's finding by comparing the effects of various durations of signaled and unsignaled delay. To assess the generality of any effect, the present experiment also included two different types of reinforcement schedules. Variable-interval 60-sec and differential-reinforcement-oflow-response-rates (DRL) 20-sec schedules were selected because it seemed possible that delayed reinforcement might produce similar effects on the behavior maintained by these 145 1981, 35,[145][146][147][148][149][150][151][152] NUMBER 2 (MARCH)