In a series of three experiments, forty Sprague-Dawley females were randomly assigned to groups which differed in incentive magnitude in the goalbox, The subjects were then trained for 10 days in a straight alley with no obstruction of approach to the goal. During testing, the rats received two nonblocked and four blocked (delay) trials per day for 11 days. Groups receiving four or nine 45-mg pellets on each trial ran significantly faster following delays than following no delay and tended to be faster following a 4-sec delay than following longer delays. Delay had a similar effect on both running speeds and ingestion rates, with the 4-and 20-sec delays producing a significant frustration effect for running and ingestion in the third experiment.The frustration effect as originally reported by Arnsel and Roussel (1952) was demonstrated with rats in a double alley apparatus consisting of a startbox (SB), first alley (A,), first goalbox (G,), second alley (A 2 ) , and second goalbox (G 2 ) . The frustration effect consisted of faster running in A 2 when the rat was not reinforced in G,. Scull (1973) has recently summarized the results of much of the research involving frustrative nonreinforcement.A manipulation to which Brown and Farber (1951) attributed frustration eliciting properties, namely, introduction of a delay between the initiation and completion of a response sequence, has received very little attention. This neglect is especially perplexing since. frustration by delay appears to be confounded with nonreinforcement in the Arnsel and Roussel study and in subsequent studies using their paradigm. In short, is Amsel's "frustration effect" due to removal of an anticipated reward in G" to delaying the approach to the reward in G 2 , or to a combination of both factors?The first study specifically designed to investigate the effects of varied intervals of delay in G, on A 2 running speeds of rats which were never reinforced in G, was conducted by Holder, Marx, Holder, and Collier (1957). Both the IS-sec and the 45-sec delay groups ran faster in A 2 than the l-sec delay group. Other investigators, on the other hand, have generally failed to find a frustration effect with a delay procedure (Ludvigson, 1968;Williams & Ellis, 1970;Wist, 1962). However, Uyeno (1965), who criticized Wist for using a between groups design, found a frustration effect with a I-sec delay when using a within subjects but not when using a between groups design. It should be emphasized that This study was based on a doctoral dissertation submitted to Michigan State University in partial fulfillment of the requirements for the PhD degree. The author would like to give special thanks to his dissertation director, M. Ray Denny, and to Mark Rilling, Stanley Ratner, and Lawrence O'Kelly for serving as committee members. Thanks also go to Mark Owen, Marcia Whiting, and Gloria Evert. Requests for reprints should be sent to John L. Allen, Spring Arbor College, Spring Arbor, Michigan 49283.The author recognizes that Experiments II and III are not ...