A RESEARCH METHODOLOGY HAS BEEN DESIGNED TO PREDICT DROPOUT BEHAVIOR AMONG URBAN NEGRO BOYS WHO CONFRONT CONVENTIONAL HIGH SCHOOL PRESSURES. IT WAS HYPOTHESIZED THAT THE DETERMINING INDICATORS OF DROPOUT BEHAVIOR LIE WITHIN THE SOCIAL-PERSONAL CHARACTERISTICS OF THE INDIVIDUAL RATHER THAN IN THE SYSTEMATIZED, SOMETIMES HOSTILE DEMANDS IMPOSED UPON THE INDIVIDUAL STUDENT BY THE SCHOOL TO WHICH HE CANNOT ADEQUATELY RESPOND OR ADJUST. THUS THE CHILD'S PRIOR PREPARATION FOR THE SCHOOL EXPERIENCE IS OF KEY IMPORTANCE. ATTENDANCE, GRADES, NUMBER OF RETAINMENTS, AND PARTICIPATION IN SCHOOL ACTIVITIES WERE IDENTIFIED AS THE CHARACTERISTIC ELEMENTS OF DROPOUT BEHAVIOR, OR "ATTENDANCE STATUS." IN TESTING THE VALIDITY OF THE HYPOTHESIZED MODEL, RESEARCH PERSONNEL FOUND THAT IQ, ACADEMIC TOOLS AND READINESS, SELF-IMAGE, PEER ADJUSTMENT, FAMILY AND COMMUNITY STATUS, FAMILY STRUCTURE, AND PARENTAL INVOLVEMENT WERE THE SOCIAL-CULTURAL VARIABLES AFFECTING ATTENDANCE STATUS. SIGNIFICANTLY, DIFFERENTIAL RESPONSES TO A COMMON SCHOOL ENVIRONMENT WERE ELICITED FROM SOCIOECONOMICALLY SIMILAR SUBJECTS. THUS, CONTRARY TO MOST SOCIOLOGICAL THOUGHT, SIMILAR ENVIRONMENTS DO NOT NECESSARILY CAUSE SIMILAR ACADEMIC PERFORMANCE AND INTRA-CLASS ACADEMIC DIFFERENCES DO EXIST. IN AN ATTEMPT TO TEST THE RELIABILITY OF THE PREDICTIVE INSTRUMENTS, IT WAS FOUND THAT SCHOOL RECORDS AND FAMILY STRUCTURE VARIABLES APPARENTLY PREDICT EARLY SCHOOL LEAVERS BETTER, AND THAT THE PARENTAL INVOLVEMENT AND PERSONAL-SOCIAL RELATIONS VARIABLES APPEAR TO PREDICT THE LATE LEAVERS BETTER. THESE FINDINGS MIGHT BE USED TO ESTABLISH CLINICAL METHODS OF DROPOUT COUNSELING. RELIABILITY TESTING NEEDS TO BE CONTINUED. TABLES ARE APPENDED. (LB)