The effect of programmed tutoring on mathematics achievement for kindergarten and first grade children was investigated. The program used discovery-based activities, with each child working on a one-to-one basis with an adult aide. The adult tutor received about 30 hours of training in the program but no direct instruction in mathematics or classroom teaching methods. Significant differences were found in concepts and computational skills when the "Programed Math Tutorial" was used as a supplement to regular instruction. A listing of the eleven major content topics in the program is included. (LS)