AN INSTRUMENT FOR MEASURING VISUAL DISCRIMINATION ABILITY WITHOUT CONFOUNDING VARIABLES OF MOTOR SKILLS HAS BEEN CONSTRUCTED. TO AVOID THE EYE-HAND COORDINATION USUALLY REQUIRED ON THESE TESTS, A SELECTION, RATHER THAN A DRAWING, RESPONSE WAS DEVELOPED. THIS TEST, THE UNIVERSITY OF CALIFORNIA AT LOS ANGELES DISCRIMINATION INVENTORY (VDI), CONSISTS OF 52 ITEMS ON FOUR SUBTESTS--FIGURE GROUND, FORM CONSTANCY, CLOSURE, AND POSITION IN SPACE. EACH OF THESE HAS BEEN ASSUMED TO REPRESENT PREREQUISITY SKILLS IMPORTANT IN LEARNING TO READ. CHILDREN THREE, FOUR, AND FIVE YEARS OLD WERE TESTED ON THE VDI AND TWO MEASURES OF MENTAL ABILITY. RESULTS INDICATE THAT SIGNIFICANT DIFFERENCES ARE ATTRIBUTABLE ONLY TO AGE AND RACE. CORRELATIONS BETWEEN THE VDI AND INTELLIGENCE MEASURES ARE WEAK. ALTHOUGH RELIABILITY, FACE VALIDITY, AND CONSTRUCT VALIDITY HAVE BEEN ESTABLISHED, A STUDY IS NOW BEING PLANNED TO MEASURE PREDICTIVE VALIDITY. ADDITIONAL WORK IS ALSO BEING DONE IN DEVELOPING THE FOUR SUBTESTS AS MEASURES OF COMPONENT ELEMENTS OF A BROAD VISUAL DISCRIMINATION ABILITY. IT IS EXPECTED THAT THE FINAL VDI WILL BE USED WITH CHILDREN FROM A WIDE RANGE OF SOCIOECONOMIC BACKGROUNDS AND WILL IDENTIFY THOSE CHILDREN WITH A VARIETY OF LEARNING DISORDERS, PARTICULARLY DYSLEXIA, BEFORE THEY BECOME SEVERE ENOUGH TO ALERT CLASSROOM TEACHERS. (AUTHORS/PR)