A copying test (VAT) is described in which the test items can function as teaching objectives with the expectation that acquiring competency in the represented behaviors will be generalized to other visual-motor tasks. Six hundred and sixty-seven kindergarten first- and second-grade children were tested. Results show: (1) a reliable scoring method (Interrater r=.98); (2) a broad range of individual differences between and within the three age groups; (3) significant predictive validity to the norm-referenced Rutgers Drawing Test (K and Gr. 1: VAT-Rutgers Form A, r=.80; Gr. 2: VAT-Rutgers Form B, r=.68). Item difficulty is analyzed and implications are discussed. (Author)