Eighty-four nine year old children, twelve in each of seven experimental groups, learned a two-choice successive discrimination problem. The parameters that were systematically manipulated included: (1) immediate versus delayed reinforcement; and (2) forced preresponse stimulus exposure versus stimulus exposure during the delay interval, intertrial and interstimulus intervals. The typical decrease in performance found under delayed reinforcement was obtained only when: (1) no preresponse exposure was required; and (2) when the discriminative stimuli were absent from Ss view during the delay of reinforcement period. Results are discussed within the framework of perceptual learning theory. (Author)