The effect of stimulus novelty, an attentional variable, on learning set acquisition was investigated. Learning set (LS) acquisition refers to an improvement in performance across a series of problems which have a common basis of solution. The design of this study involved two groups, one in which the positive stimulus on Trial 2 involved the introduction of a novel dimension (Novelty-Positive), and one in which the negative stimulus involved novelty (Novelty-Negative). The novel dimension introduced was color, and was preceded by both rewarded and nonrewarded first trials. Twenty-four 3-year-old children were randomly assigned to one of the two experimental treatments. The results supported the novelty hypothesis in that performance was significantly better in group Novelty-Positive than in group Novelty-Negative. No significant reward differential was found in this experiment as has been found in other studies (e.g., Moss & Harlow, 1947). The result of the present experiment replicates a similar finding reported by Callaghan and Cameron (1978). (Author/SS)