Three broad categories of conditions, with 12 particular subconditions, have been hypothesized as specifiers of the SES/delinquency relationship. Here, we review the recent empirical literature as it bears on these potential specifications. The results do not support any of the conditional hypotheses about SES and delinquency, and they again challenge the idea that a negative SES/delinquency relationship is general and pervasive. Almost all of the recent research finds some condition under which SES and delinquency are significantly related, however, and several of the specification hypotheses have not been thoroughly enough investigated to permit firm conclusions about their potency. This poses a quandary for scholars trying to understand delinquent behavior. Possible responses to the situation are discussed. 55 references. (Author abstract)