The questions in this study were three-fold. The first question, suggested by the work of Das and Naglieri (e.g., Naglieri & Das 1987, 1988), addressed the hypothesis that a higher-order structure underlies performance on a battery of cognitive tests. The second question, concerned with external validity, addressed how well the higher-order cognitive ability factors predict two criterion-related measures of ability. Academic Achievement and Word Skills. The third question, also concerned with external validity, addressed whether differences between students classified as low or normal in reading achievement were mediated by individual differences in the higher-order ability factors. That is, do the hypothesized higher-order cognitive abilities play a mediating role in predicting deficits in academic achievement and reading competence? Structural equation modeling techniques were used to test these research hypotheses on a stratified random sample of 135 third grade students. The first stratum contained a random selection of 69 students classified as low in reading achievement based on the Chapter 1 entitlement assignments made by the school district the year prior to the study and the second stratum contained 66 students selected from the normal achieving population. The results showed that (a) two higher-order factors, Planning/Attention and Successive Processing, explained the relations among six lower-order ability constructs, (b) the higherorder ability factors predicted individual differences in the two criterion measures of Academic Achievement and Word Skills, and, (c) low achieving students showed significant deficits only in the two higher-order factors, suggesting a mediating role for the cognitive skills represented by the higher-order factors.