This investigation attempted to clarify the effects of different classes of reinforcement on the size discrimination learning of normal and low achieving third grade middle class boys. The classes of reinforcement consisted of tangible reinforcement, person reinforcement (praise), and performance reinforcement (correctness of response). Each of these reinforcers was presented on a reward-nothing combination basis. The results indicated that the subjects did not all respond homogeneously to a particular class of reinforcement when achievement level was considered. The low achieving boys had a higher percentage of correct responses on the learning task under tangible reinforcement than under person or performance reinforcement, while the normal achieving boys performed more effectively under person and performance reinforcement than under tangible reinforcement. (Author/TA)