Public Law 94-142 mandates the identification and placement of learning disabled children based primarily on the measurement of intelligence. It is, therefore, the responsibility of educational psychologists to use standardized intelligence tests appropriately to accurately and objectively assess a childs' intellectual potential and ability. Three general assumptions underlying the measurement of intelligence are discussed: (1) that intelligence is both measurable and quantifiable; (2) that it is distributed according to a normal curve; and (3) that intelligence remains constant over time. The constancy of intelligence level and the stability of its measurement (test reliability) are important issues. Test reliability may be established according to internal consistency; correlation with equivalent or parallel tests; or, most importantly, consistency of measurement over time. Data on the stability of certain intelligence tests over time are presented. These limited data suggest that the scores of young children are less stable than of older children and adults; that stability decreases as the length of the test-retest interval increases; and that children with various disabilities exhibit more test-retest variation. A formula is presented for estimating an unbiased true score. More research is needed investigating fair methods of placing learning disabled children. (GDC)