The application of the labeling perspective in relation to juvenile delinquency is examined. The general theoretical origin of labeling theory in both conflict theory and symbolic interactionism is discussed and used to present two main labeling hypotheses: 1) that status attributes are influential in determining who is labeled (the "status characteristics hypotheses") and 2) that labeling experiences are instrumental in producing problems of adjustment and in causing subsequent commitment to further deviance (the "secondary deviance hypotheses"). It is noted that what is often passed off as a critique of labeling theory itself is frequently a caricature of the theory. The major point of the paper is an elaboration of the full complexity of each labeling hypothesis, a suggestion for empirical research to test it, and a review of the extant literature. It is suggested that labeling theory is not as invalid as its critics have claimed, and that what is needed is a restatement and revitalization of a labeling theory of delinquency. Over 150 references. (Publisher abstract modified)