TWO GROUPS OF CHILDREN ATTENDED THE UNIVERSITY OF KANSAS NURSERY SCHOOL FOR EIGHT WEEKS. THE MORNING CLASS WAS HELD FOR 20 FOUR-YEAR-OLD CHILDREN OF MIDDLE-CLASS FAMILIES. THE AFTERNOON CLASS WAS FOR 24 FIVE-YEAR-OLD CHILDREN OF LOW-INCOME FAMILIES. THREE COMPARATIVE STUDIES WERE MADE. IN STUDY I NO CHANGE WAS FOUND ON THE PEABODY PICTURE VOCABULARY TEST IN THE MIDDLE-CLASS GROUP, WHEREAS AND INCREASE IN SCORES WAS FOUND FOR THE HEAD START GROUP. THE PRESCHOOL INVENTORY SHOWED THE MIDDLE-CLASS PRESCHOOL GROUP TO BE SIGNIFICANTLY HIGHER IN PERFORMANCE, BUT SOME CHANGES DID OCCUR OVER THE SUMMER FOR THE HEAD START GROUP. IN STUDY II NO OVERALL DIFFERENCES WERE FOUND BETWEEN THE GROUPS ON TWO DISCRIMINATION LEARNING TASKS, BUT AN INTERACTION OF GROUP AND SEX WAS INDICATED. IN STUDY III CHILDREN JUDGED TO BE LOW IN SOCIAL RESPONSIVENESS WERE SELECTED FROM THE TWO GROUPS, AND THEIR PERFORMANCE IN A LABORATORY SETTING WAS ASSESSED. THE MAJOR DIFFERENCES BETWEEN THE HEAD START AND MIDDLE-CLASS GROUP IN A BASELINE ASSESSMENT APPEARED TO BE IN VOCALIZATIONS. A FINAL IMPLICATION OF THIS STUDY IS THAT THE PERSISTENT BEHAVIORAL DEFICIENCES OF THE HEAD START CHILDREN REVEALED IN THE BASELINE SESSIONS APPEAR TO BE REVERSIBLE. (CO'D)