This study investigated the construct validity of five perceptual traits (auditory discrimination, visual discrimination, visual memory, visual-motor coordination, and auditory to visual-motor coordination) with five simulated work samples (union assembly, resistor reading, budgette assembly, lock assembly, and nail and screw sort) from the Jewish Employment Vocational Service System (JEVS). The subjects (mean age was 14.4 and mean intelligence quotient was 94.0) were 36 learning disabled applicants attending a prevocational assessment program in Monmouth County, New Jersey. Five subtests, from three different tests, were administered to the subjects. The tests included: (1) the Weschler Intelligence Scale for Children-Revised; (2) the Specific Language Disability Test, and (3) the JEVS. The test scores were analyzed by Pearson product-moment correlation. Utilizing the paradigm of convergent-discriminant validity, convergent validity was evidenced in four of the five work samples and discriminant validity was demonstrated in three of the five work samples. The following work samples measured a distinct perceptual construct: (1) union assembly measures visual memory; (2) budgette assembly measures visual discrimination; and (3) lock assembly measures auditory discrimination. The results indicate that simulated work samples can be used as a diagnostic instrument for assessment of perceptual abilities in a vocational setting. (PN)