Using Festinger's theory of cognitive dissonance as a model, this study attempted to change the attitude and behavior of children toward well liked toys. The results offer only limited support for the theory. The subjects in the three groups did play a significantly different amount of time in the two play periods. The t-tests indicated it was the children who received the mild threat who were playing differentially. They played less in the Post play period and most in the Final play period. This study does not support Freedman's (1965) findings of the effects being maintained across time. These results indicate that the effects of not playing initially are at least partially compensated for later. The results of this study though tending to support the theory of cognitive dissonance for short-term behavior change raise doubts about generalizing the positive results of published studies to populations that have not been investigated. It further indicates that more careful empirical study should be given the nature of threats and toys selected, for use in studies of this kind. Finally, this study warrants the conclusion that studies of dimensions as complex as attempted attitude and behavior change in young children require the utmost precision and preliminary research to rule out other factors which may effect results in an unascertained manner. (Author/KJ)