The power of the Slingerland Screening Tests was examined, and a discrepant performance index was developed between cognitive measures to predict early school achievement and class placement for children at risk for specific language disabilities. Subjects were two successive cohorts of entering kindergarten children for whom Slingerland Tests scores and scores from either the Cognitive Abilities Test or the Raven Coloured Progressive Matrices were obtained. Correlational analysis was performed using these scores, subsequent kindergarten achievement and first-grade class placement for both cohorts, and first-grade achievement for the first cohort. Modest predictive validity for the Slingerland Tests was observed, stronger than the Raven Test and weaker than the Cognitive Ability Test for overall achievement. Weak but significant class placement prediction was also observed for the Slingerland Tests. An index based upon discrepant performance on the Slingerland and the more standard measures of mental ability produced inconsistent results and was redundant with single measures of general ability used alone. (Author/GDC)