THE FRAMEWORK, CALLED A COHORT MODEL, MAKES IT POSSIBLE TO ORGANIZE AND CODIFY INFORMATION ABOUT DRUG USE AND DRUG ATTITUDES WITHIN MANY SUCCESSIVE COHORTS, E.G., HIGH SCHOOL STUDENTS, COLLEGE STUDENTS, AND VIETNAM VETERANS. THE PEOPLE WITHIN EACH COHORT ARE VIEWED AS AGING, I.E., PASSING THROUGH A SEQUENCE OF ROLES (DEPENDENT CHILD, STUDENT, WORKER, ETC.) FROM BIRTH TO DEATH. THEY LEARN TO PLAY NEW ROLES AND RELINQUISH OLD ONES. THEY ACCUMULATE KNOWLEDGE, ATTITUDES, AND SOCIAL EXPERIENCES AND UNDERGO BIOLOGICAL AND PSYCHOLOGICAL DEVELOPMENT AND CHANGE. AS PARTICULAR INDIVIDUALS ARE AGING AND DYING, NEW COHORTS ARE CONTINUALLY BEING BORN. MEANWHILE THE SOCIAL STRUCTURE CONSTANTLY CHANGES AS SOCIETY UNDERGOES WARS, FAMINES, PERIODS OF PROSPERITY AND DEPRESSION, CHANGES IN THE STATE OF SCIENCE AND THE ARTS, AND REVOLUTIONS IN TASTES AND BELIEFS. AS SOCIETY CHANGES, EACH NEW COHORT ENCOUNTERS A UNIQUE SEQUENCE OF SOCAL AND ENVIRONMENTAL EVENTS. THUS, THE LIFE-COURSE PATTERNS OF PEOPLE IN ONE COHORT DIFFER IN SOME RESPECTS FROM THE PATTERNS OF THOSE IN OTHER COHORTS. DIFFERENT COHORTS AGE IN DIFFERENT WAYS, AND THESE DIFFERENCES CANNOT BE OBSERVED FROM LONGITUDINAL STUDIES OF SINGLE COHORTS. COHORT COMPARISONS PERMIT THE IDENTIFICATION OF SIMILARITIES AND DIFFERENCES AMONG LIFE-COURSE PATTERNS OF DIFFERENT COHORTS AND LEAD TO AN UNDERSTANDING OF HOW THE CONSECUTIVE COHORTS, EACH LAGGING BEHIND ITS PREDECESSOR IN THE PROCESS OF AGING, ALL FIT TOGETHER TO FORM THE CHANGING AGE STRUCTURE OF SOCIETY. IN ADDITION TO SUPPLEMENTING COHORT-SPECIFIC LONGITUDINAL RESEARCH WITH CROSS-COHORT COMPARISONS, IT IS NECESSARY TO CONSIDER THE COEXISTENCE OF MANY COHORTS WHO, DESPITE THEIR DIFFERENCES IN AGE AND HISTORICAL EXPERIENCE, SHARE THE SAME SOCIETY AND POSSIBLY EXERT PROFOUND INFLUENCES ON ONE ANOTHER. IN LIGHT OF THESE CONCEPTS, DRUG ABUSE RESEARCHERS ARE ALSO URGED TO EMPLOY A WIDE RANGE OF STRATEGIES, COMPLEMENTING LONGITUDINAL DESIGNS WITH CROSS-SECTIONAL SURVEYS. RESEARCHERS ARE ALSO URGED TO COORDINATE THEIR OBJECTIVES, STANDARDIZE KEY DEFINITIONS, AND CONSIDER THE COMPLEMENTARITY OF THEIR RESEARCH DESIGNS, SO THAT ACCUMULATION AND CODIFICATION OF KNOWLEDGE ON DRUG ABUSE WILL BE POSSIBLE. EXAMPLES OF THE USE OF COHORT COMPARISONS IN STUDIES OF DRUG USE INITIATION, LIFE-COURSE CHANGE, AND THE ANTECEDENTS AND CONSEQUENCES OF DRUG USE ARE CITED. A LIST OF REFERENCES IS INCLUDED. (AUTHOR ABSTRACT MODIFIED--LKM)