Questions are addressed pertaining to the Language Assessment Battery (LAB), currently the only single battery that meets all of the state of Connecticut criteria for an instrument to determine which pupils must be offered bilingual education. Normed on an English dominant population, the primary purpose was to obtain its psychometric characteristics for pupils of limited English proficiency. Other purposes were to determine its selection accuracy, and to examine it as a measure of English achievement within a bilingual program. Validity was also reviewed. The psychometric portion of the study used 195 K-2 Hispanics in a bilingual program. Scale structure, internal consistency reliability, and item statistics were analyzed. The other investigations used the publisher's information. The problems encountered were instructive for the makers of this type of test. Psychometrically the test is sound except for the low reliabilities and poor item statistics of the reading and writing subscales in K and grade 1. Those scales at those grades also have little content validity. Standard error of measurement is very poor in terms of percentile rank near the cut scores used for selection. Establishment of standards of construct validity for this new type of test are recommended. (Author/GK)