With few exceptions, virtually all research on the etiology of adolescent substance abuse has focused entirely on identifying risk factors while ignoring protective factors that insulate teenagers from drug involvement. This perspective fails to consider that protective forces may operate in both direct and interactive ways. This longitudinal study tested new approaches that consider a wide range of factors in indices of risk, protection, and their interaction. Analyses are based on two waves of data collected in a 12-year longitudinal study of adolescent and adult development and drug use. In year 5, data were collected from 896 late adolescents (grades 11, 12, and post high school). Multiple items of three specific types of drug problems were used to reflect general latent constructs of alcohol problems, cannabis problems, and cocaine problems. The three specific types of problems were assessed by the frequency of occurrence of various adverse consequences attributable to drug use. Bivariate, multiple regression, and latent-variable structural equation analyses were used to examine the relationships between vulnerability (risk, protection, and their interaction), and frequency and quantity of drug use (cigarettes, alcohol, cannabis, cocaine, and hard drugs), and between vulnerability and problems related to drug use 8 years later. Vulnerability in adolescence was found to have both a direct and indirect influence, and lead to problem use of cannabis, alcohol, and cocaine in adulthood. (LLL)