Each year large numbers of children experience academic failure during the first 3 years of their public school involvement. Many studies have been done using demographic and norm-referenced tests to try to identify at-risk-for-failure youngsters either at the preschool level or early during their school experience. Most attempts have been costly and of limited success. This study examined the concurrent and predictive validity of performance on a test measuring the information processing characteristics of students to identify youngsters with low achievement in school. Second-grade students (N=380) were individually administered the Learning Efficiency Test (LET) and regression and discriminant analyses were conducted. Variables examined included actual levels of performance in reading and mathematics, California Achievement Test total scores in reading and mathematics, and teacher rankings of overall student achievement levels. Results showed high levels of accuracy for the LET in predicting both achievement levels and in classifying students according to teacher rankings. The LET is easy to administer and reflects a cost-effective approach. Moreover, the predictive efficacy of the LET represents substantial improvements over past studies that used costly and elaborate diagnostic procedures which have failed to identify at-risk students beyond chance levels. (Author/ABL)