Duncan et al. (2007) examined associations between early behavioral and cognitive skills with later achievement. These associations were examined in 6 different data sets and results converged to suggest that early behavioral competences or problems had little, if any, prediction to later achievement and that attentional competences had small positive relations with later achievement. In contrast, cognitive abilities were by far the strongest predictors of achievement. We provide and investigate potential reasons why Duncan et al. found little to no association between behavior and later achievement in a reanalysis of data from 3 studies previously analyzed by Duncan et al. Potential reasons include the validity of the behavioral measures, treatment of the behavioral measures as continuous as opposed to categorical, and the choice of data analytic method. In this article, we discuss these issues at greater length and address them in our reanalysis. We also bring into question the nature of the relationship between behavior and achievement. Generally, our reanalysis supports the idea that attention measures are more predictive than behavioral measures; however, certain behavior measures showed small to moderate associations to concurrent levels of academic achievement and changes in academic achievement through elementary school. (Contains 3 tables.)