This paper reports on one study in a research program trying to extend existing findings on elaborative mediation in paired-associate learning into the domain of prose learning. Ten third graders and 22 fourth graders from an inner-city Catholic school served as subjects. A paraphrase recall pretest was administered which involved reading a three-paragraph, 176-word passage and then retelling it, not necessarily verbatim, from memory. The two groups then received twelve training sessions over a period of 27 days. Two days after the final training session, subjects were given a paraphrase recall test without any imagery prompting, a paraphrase recall test with imagery prompting, and Form G of the Metropolitan Achievement Test (MAT) Elementary Reading battery. The results indicated that there were no differences between groups except on the paraphrase recall test with imagery instruction, where the experimental scores were forty greater than the control scores. (WR)