A 1978 study (Torgesen, Bowen and Ivey) of the structure and modality variables of the Visual-Aural Digit Span (VADS) test was replicated to determine: (1) if the effects generalized across age; (2) if differences between simultaneous and sequential visually presented items were due to mode of presentation or the amount of study time; (3) the effects of mode of output; and (4) the relationship between performance on the various tests and performance on standard achievement tests in mathematics and reading for second and fourth graders. The major consistent finding was that all groups performed significantly better on simultaneous digit span tests than on sequentially presented digit span tests. However, there was no significant difference in overall performance between second and fourth graders. Unlike Torgesen et al., the consistent positive and significant correlations among performance on all conditions argued for a consistent memory for order across different conditions. Finally, the failure to find significant correlations between the memory for order measures and achievement scores fails to replicate earlier studies. Possible explanations for the failure to replicate revolves around differences in age of subjects and measures used. (Author/PN)