In recent years, considerable attention has been given to a social information processing model as a means of understanding interaction patterns in children. Within the framework of that model, systematic biases have been found in the manner in which aggressive children process social information. The present study sought to extend that literature by applying the model to younger children, by examining the processing of withdrawn, as well as aggressive, children, and by employing traditional affect recognition tasks as the stimuli. Sixty preschool children were nominated by their classroom teachers as either aggressive, withdrawn, or well adjusted, according to their predominant interaction style. The children were then tested, using a set of affect recognition tasks which assessed stimulus encoding and interpretation. Stimuli consisted of facial expression photos and context stories portraying one of four emotions (Happy, Sad, Mad, or Neutral). The hypotheses of the study predicted systematic biases in stimulus encoding and interpretation, consistent with the subjects’ behavioral style. Analyses failed to support the hypotheses in that the groups failed to show identifiable systematic biases. Exploratory analyses revealed that subgroups of subjects demonstrated such biases, but those biases were related only to level of developmental maturity. The discussion of the findings explored issues which may have led to the negative results. Further research directions were also discussed which will help to clarify the questions raised by the present study.Ph. D.