The predictive power of the Split-Half version of the Wechsler Intelligence Scale for Children--Revised (WISC-R) Object Assembly (OA) subtest was compared to that of the full administration of the OA subtest. A cohort of 218 male and 49 female adolescent offenders detained in a Texas juvenile detention facility between 1990 and 1992 was used. The correlation between the scores obtained on the OA subtest in the Split-Half version and the full administration indicates that the brief administration is effective and saves time. The short form may have considerable clinical utility. It is important to remember that these results do not generalize to the other nine WISC-R subtests or to the Verbal, Performance, and Full Scale intelligence quotients. (SLD)