Thirty-five second-grade learning disabled children participated in a visual information processing training program designed to teach analysis of visual material into component parts, systematic scanning of visual arrays, pick-up, description, and memory storage of distinctive information, and efficient solution of visual match-to-sample problems. After eight weeks in this experimental program, the treatment group was superior to contrast and control groups on a battery of visual information processing tasks, the Embedded Figures Test, the Matching Familiar Figures Test, and two reading achievement measures. Results from a delayed posttest indicated that the gains were maintained throughout two months of summer vacation. (Author)