A STUDY DESCRIBED AS THE FIRST APPLICATION OF CROSS-VALIDATED UNIQUENESS ANALYSIS TECHNIQUES WAS DESIGNED TO ELIMINATE THE EFFECTS OF IMPERFECTIONS IN A PRIOR FACTOR-ANALYTIC STUDY OF READING COMPREHENSION WHICH USED TESTS ESPECIALLY CONSTRUCTED TO MEASURE MENTAL SKILLS IN READING. A UNIQUENESS ANALYSIS BASED ON LARGE SAMPLES WAS USED TO OBTAIN ESTIMATES OF THE PERCENTAGE OF NONCHANCE VARIANCE. STUDENTS IN GRADE 12 NEAR PHILADELPHIA SERVED AS SUBJECTS FOR THE PRELIMINARY AND MAIN STUDIES. EIGHT READING COMPREHENSION SKILLS WERE TESTED. IN THE PRELIMINARY STUDY, TWO PARALLEL FORMS OF THE TEST WERE ADMINISTERED TO APPROXIMATELY 400 STUDENTS. A POOL OF 24 ITEMS FOR EACH OF THE EIGHT SKILLS WAS ASSIGNED TO TWO PARALLEL FORMS OF THE TEST. APPROXIMATELY 1,000 STUDENTS TOOK BOTH FORMS OF THE TEST WITH 1 OR 2 DAYS INTERVENING. INTERCORRELATIONS AND MULTIPLE REGRESSION ANALYSES WERE USED TO ANALYZE THE DATA. IT IS CONCLUDED THAT COMPREHENSION AMONG MATURE READERS IS NOT A UNITARY MENTAL SKILL OR OPERATION. IT IS POINTED OUT THAT SYSTEMATIC LEARNING EXERCISES APPROPRIATE IN LEVEL OF DIFFICULTY FOR EACH PUPIL SHOULD BE PROVIDED. IN ADDITION, THERE IS A NEED FOR A SERIES OF SELF-TEACHING PRACTICE EXERCISES FOR DEVELOPING PROFICIENCY IN THE CONSTITUENT SKILLS OF COMPREHENSION AMONG MATURE READERS. A BIBLIOGRAPHY IS GIVEN. (BK)