THE TWO MAJOR PURPOSES OF THE STUDY WERE (1) TO DETERMINE THE RELATION OF ANXIETY ABOUT SCHOOL TO A WIDE VARIETY OF PERSONALITY AND BEHAVIOR VARIABLES MEASURED IN THE CLASSROOM AND (2) TO EXPLORE THE EFFECTS OF DIFFERENCES IN A CLASSROOM ATMOSPHERE UPON THOSE RELATIONS. THE TEST ANXIETY SCALE FOR CHILDREN (TASC) AND THE DEFENSIVENESS SCALE FOR CHILDREN (DSC) WERE THE TECHNIQUES UTILIZED TO GET CERTAIN SELF-ATTITUDES. ALL THE CHILDREN OF THREE SECOND-GRADE CLASSROOMS IN THE SAME ELEMENTARY SCHOOL WERE THE SUBJECTS. THEY WERE ADMINISTERED THE TASC, DSC, AND LORGE-THORNDIKE TESTS AFTER THE FIRST GRADE AND OBSERVED BY TWO DIFFERENT OBSERVERS OVER A 4-MONTH PERIOD IN THE SECOND GRADE, AND RATED ON A PERSONALITY BEHAVIOR CHECKLIST. THE SAME CHILDREN WERE STUDIED IN THE THIRD GRADE TO DETERMINE THE CONSISTENCY OF THE FINDINGS AND THE EFFECTIVENESS OF THE TASC AND DSC AS RELATIVELY LONG-RANGE PREDICTIONS. ALSO, THE THIRD-GRADE TEACHERS RATED THE PUPILS ON THE PERSONALITY AND BEHAVIOR CHECKLIST. THE MAJOR FINDINGS WERE--(1) THE DSC IS A BETTER PREDICTOR THAN THE TASC OF CHILDREN'S PERSONALITY AND BEHAVIOR IN THE CLASSROOM, (2) THE DSC PREDICTS BETTER FOR GIRLS, TASC FOR BOYS. DSC IS A MORE STABLE MEASURE REGARDLESS OF CHANGE IN TEACHERS, AND (3) TEACHER-OBSERVER AGREEMENT ABOUT CLASSROOM BEHAVIOR TENDS TO BE HIGHEST FOR THOSE TEACHERS INDEPENDENTLY RATED AS MOST EFFECTIVE. IN GENERAL, TEACHERS' VALUES AND BELIEFS AFFECT THEIR OWN AND THEIR PUPILS' CLASSROOM EFFORTS, AND IN PARTICULAR THESE AFFECT CHILDREN'S ANXIETY AND DEFENSIVENESS, AND BOYS AND GIRLS DIFFERENTLY. (GC)