Score linkages between the Verbal and Math sections of the SAT I: Reasoning Test and the corresponding sections of the new version of a Spanish-language admissions test, the Prueba de Aptitud Academica (PAA), were investigated. A bilingual group design was employed. A language proficiency measure (ESLAT) was used to define the bilingual group and as a predictor variable. Prediction and scaling for concordance results were compared. Results indicated that for both single (PAA Verbal or PAA Math to the corresponding SAT I scores) and composite (PAA-V+M to SAT I-V+M and PAA-V+M+ESLAT to SAT I-V+M) score linkage, prediction is preferable to concordance. Comparison of prediction and concordance results for composite scores versus single construct scores indicates that when PAA Verbal is combined with PAA Math to form a composite, predictions of this composite are better than for Verbal alone but worse than for Math alone.