THE QUESTION OF PRIMARY CONCERN IN FOUR STUDIES WAS IMMEDIATE FEEDBACK IN ADJUNCT AUTOINSTRUCTIONAL EXERCISES (LEARNING EXERCISES DEFINED AS THOSE DESIGNED TO TAKE ADVANTAGE OF THE STRUCTURE INHERENT IN SUBJECT MATTER). IN A CONTROLLED SITUATION, DIFFERENT FEEDBACK DELAYS (IMMEDIATE, 1-HOUR DELAY, 2-DAY DELAY, AND 1-WEEK DELAY) WERE TESTED. OTHER OBJECTIVES WERE (1) TO PREPARE AND TEST ADJUNCT AUTOINSTRUCTIONAL EXERCISES IN FOUR COURSES REPRESENTING DIFFERENT SUBJECT-MATTER (COMPUTER TECHNIQUES, HIGH SCHOOL ALGEBRA, JUNIOR HIGH SCHOOL GENERAL SCIENCE, AND EDUCATIONAL PSYCHOLOGY) AND AT DIFFERENT GRADE LEVELS, AND (2) TO DETERMINE WHETHER CLASSROOM TEACHERS WITH SUFFICIENT ENCOURAGEMENT CAN DEVELOP ADJUNCT AUTOINSTRUCTIONAL EXERCISES. MATERIALS WERE DEVELOPED FOR USE IN COURSES IN THE FOUR SUBJECT MATTERS. THE WORK WITH THESE COURSES CONSTITUTED FOUR SEPARATE EXPERIMENTAL STUDIES, REPORTED SEPARATELY BECAUSE OF VARIATIONS IN PROCEDURES. SIGNIFICANT DIFFERENCES ON FINAL EXAMINATION SCORES OF PARTICIPATING STUDENTS WERE NOT DEMONSTRATED BETWEEN EXPERIMENTAL GROUPS WHICH RECEIVED IMMEDIATE KNOWLEDGE OF RESULTS AND CONTROL GROUPS WHICH DID NOT RECEIVE SUCH FEEDBACK IN THREE SUBJECTS. SIGNIFICANT DIFFERENCES WERE FOUND ON MIDTERM EXAMINATIONS FOR COMPUTER TECHNIQUES. IN THE COURSE IN EDUCATIONAL PSYCHOLOGY THERE WERE FOUR TREATMENT CONDITIONS. IN NO INSTANCE WAS IMMEDIATE FEEDBACK SUPERIOR TO DELAYED FEEDBACK. TOO FEW TEACHERS WERE INVOLVED IN THE EXPERIMENTS TO GET A DEFINITE ANSWER RELATIVE TO TEACHERS PREPARING ADJUNCT AUTOINSTRUCTIONAL EXERCISES. (REFER TO ACCESSION NUMBER ED 003 336 FOR INFORMATION ON AN EARLIER STUDY OF ADJUNCT AUTOINSTRUCTION.) (JC)