In this study, we examine late adolescence as a particularly critical developmental period during which each of those activities is being established. Several theories are proposed and tested to explain the relationship between alcohol use and criminal involvement using one-year longitudinal data. Sensation seeking was included as a plausible underlying spurious influence that might account for the association between alcohol use and criminal behavior. A majority of both males and females had used alcohol substances, and a substantial minority had engaged in some type of other deviant behavior. All analyses were conducted separately by sex, and -- although there were several mean sex differences -- there were few major differences in the magnitude or patterns of association between alcohol and deviance either cross-sectionally or longitudinally. Alcohol use and deviance were moderately stable over the one-year period. Nevertheless, clear support was found for the hypothesis that early alcohol use generated or caused increased later deviance and criminal behavior for both sexes. Little support was found for the reverse hypothesis, particularly when sensation seeking was considered; early delinquent behavior had no direct effect on changing patterns of alcohol use. Some support for the problem behavior hypothesis was found based on the high intercorrelations among the constructs, although little support emerged for any of the other theories considered. 4 tables, 2 figures and 53 references. (Publisher abstract)