It was hypothesized that the presence of a negative set moderates the effect of a child’s observing the behavior of an adult model. Such a set presumably can have any 1 of 3 different affects. First, the set may produce inefficient attending behaviors, or second, may cause the S to delay performance until the model is absent. Finally, such a set increases the probability that the child will behave in a manner which is consistently the opposite of behaviors displayed by the model. A laboratory procedure was described for measuring neagtive set. In the first study, boys and female adult models participated in this procedure and in a series of modeling tasks. The data showed that negative set scores correlated negatively with scores assessing the effects of modeling. Modeling tasks involving the manipulation of either complex problem-solving skills or the alteration of already well-established object preferences seemed to be most affected by negative set. In a second study, using adult models, older boys were shown to have lower negative set scores than younger boys. These findings are in agreement with the literature showing a consistent decrease in “negativism” for children between the ages of 2 and 12 years.