THE OBJECTIVES OF THIS STUDY WERE (1) TO IDENTIFY AND COMPARE THREE KINDS OF TRANSFER EFFECTS--AN EFFECT ASSOCIATED WITH CUE REPETITION, A LEARNING-TO-LEARN EFFECT, AND A WARM-UP EFFECT, AND (2) TO EVALUATE THE USEFULNESS OF CUE-RESPONSE CORRELATIONS IN EXPLAINING TRANSFER EFFECTS. THE STUDY WAS BASED ON THE DEFINITION OF TRANSFER EFFECTS AS CHANGES IN ABILITY TO DEAL WITH SITUATIONS NOT ENCOUNTERED DURING TRAINING. SUBJECTS WERE UNDERGRADUATE COLLEGE STUDENTS WHO WERE DIVIDED INTO SIX GROUPS. FOR HALF OF THE SAMPLE, THE TRAINING AND CRITERIA TASKS WERE SIMILAR. FOR THE OTHER HALF, THE TRAINING AND CRITERIA TASKS WERE QUITE DIFFERENT. FOR ONE GROUP FROM EACH HALF OF THE SAMPLE, THE RELEVANT AND IRRELEVANT CUES REMAINED THE SAME FOR ALL TASKS. FOR THE SECOND GROUP FROM EACH HALF, THE RELEVANT AND IRRELEVANT CUES WERE REVERSED ON THE CRITERION TASK. FOR THE REMAINING GROUPS, COMPLETELY NEW CUES WERE INTRODUCED DURING THE CRITERION TASK. A CONTROL GROUP PERFORMED ONLY THE CRITERION TASK. THE ENTIRE EXPERIMENT WAS CARRIED OUT TWICE, ONCE USING LARGE GROUP TESTING PROCEDURES AND ONCE WITH GROUPS OF EITHER SEVEN OR 14 AT A TIME. ERROR SCORES WERE USED AS A MEASURE OF RATE OF LEARNING. THE INVESTIGATOR ATTRIBUTED THE GREATEST TRANSFER EFFECT TO WARM-UP, AND THE NEXT LARGEST TO LEARNING-TO-LEARN EFFECTS. NO SIGNIFICANT EFFECT WAS FOUND FOR CUE REPETITION. SINGLE-TRIAL, CUE-RESPONSE CORRELATIONS (CRITERIALITIES) PROVED USEFUL IN INTERPRETING THE TEST RESULTS. THEY ALSO DEMONSTRATED THE PRESENCE OF CUE-SIMILARITY EFFECTS NOT DETECTABLE IN THE ERROR SCORES. (AL)