In a study designed to assess effects of administering the Metropolitan Reading Test (MRT) in Spanish versus English, 100 Puerto Rican kindergarten pupils were randomly split into two groups. The MRT was administered in English to one group and in a Spanish translation to the other group. The group who took the Spanish version significantly surpassed the group who took the English version with respect to total score and all but the two most nonverbal subtests. These differences were particularly evident for, but were not exclusive to, the subsample from the bilingual classes. (Author/DEP)