In a 2 by 2 factorial design, 48 albino rats were trained in a straight runway to receive either a large or a small magnitude of immediate or delayed reward, after which all groups received the large reward for an additional phase of training. The groups receiving immediate reinforcement performed more rapidly than the delayed groups throughout the experiment. The effects for reward magnitude were complex, and there was some suggestion that delayed reinforcement produced a reversal of the usual superior performance for a large reward. While delayed upshifted subjects briefly overshot the performance level of large-reward controls, group differences were not reliable, and this failure to obtain a positive contrast effect did not seem to be due to a physiological limit.