TO INVESTIGATE THE RELATIONSHIP BETWEEN EMOTIONAL CLIMATE AND LEARNING, RANDOM SAMPLES OF STUDENTS IN 49 TWELFTH GRADE PHYSICS CLASSES FROM ALL PARTS OF THE COUNTRY WERE GIVEN A CLASSROOM CLIMATE QUESTIONNAIRE WHICH WAS CORRELATED WITH THE TEST ON UNDERSTANDING SCIENCE, A PHYSICS ACHIEVEMENT TEST, AND THE SEMANTIC DIFFERENTIAL FOR SCIENCE STUDENTS. A 25 PERCENT RANDOM SAMPLE OF EACH CLASS TOOK THE CLASSROOM CLIMATE QUESTIONNAIRE WHILE A 50 PERCENT RANDOM SAMPLE TOOK THE THREE TESTS. USING MULTIPLE CORRELATION AND CANONICAL CORRELATION TECHNIQUES, CLASSES WITH HIGH GAINS IN SCIENCE UNDERSTANDING WERE PERCEIVED BY THE STUDENTS AS CONTAINING MORE FRICTION, STRICT CONTROL, AND PERSONAL INTIMACY AND LESS STRATIFICATION, GOAL DIRECTION AND SUBSERVIENCE THAN CLASSES HAVING LOW GAINS. DISORGANIZATION, FORMALITY, AND SOCIAL HETEROGENEITY WERE IN DESCENDING ORDER OF IMPORTANCE, NEGATIVELY RELATED TO PHYSICS ACHIEVEMENT GAINS. IQ SCORES WERE FOUND TO HAVE LITTLE RELATIONSHIP TO THE 3 CRITERION MEASURES. LEARNING SITUATIONS WERE SEEN AS THOSE HAVING INTENSE INTERACTION BETWEEN TEACHER AND STUDENTS, WITH THE CLASS BEING WELL ORGANIZED AND CONTROLLED BY THE TEACHER BUT WHERE THE STUDENTS WERE FREE TO QUESTION AND LEARN IN A RELATIVELY INFORMAL ATMOSPHERE. (AF)