In the past, sociologists either dismissed ot ignored the well-establi
shed empirical relationship between IQ and juvenile delinquency. More
recently, however, they have come to accept the IQ/JD connection but n
ow debate the most valid explanation for it. Some have advanced what w
e characterized as a ''school performance'' model, while others adhere
to what may be called a ''school reaction'' model. The major bone of
contention between the competing models is over the school variables t
hought to mediate the IQ/JD relationship. In this study, we specified
the most dominant versions of the competing models as well as a model
integrating the concepts of each. Using three waves of data from the Y
outh in Transition project, we first estimated parameters for each of
the competing models and then parameters for the integrated model, Con
sidered independently, each model was supported by the data. However,
when each of the separate models was forced to take into account the e
xplanatory contribution of its competitor, only the school performance
model was supported. Implications of the findings for schooling and d
elinquency are discussed.