Fifty-two low performing schools were randomly assigned to receive Spatial-Temporal (ST) Math, a supplemental mathematics software and instructional program, in second/third or fourth/fifth grades or to a business-as-usual control. Analyses reveal a negligible effect of ST Math on mathematics scores, which did not differ significantly across subgroups defined by prior math proficiency and English Language Learner status. Two years of program treatment produced a nonsignificant effect. Publication of evaluation results from large-scale real-world supplemental mathematics instructional implementations such as this one can provide a realistic view of the possibilities, costs, and limitations of this and other computer aided instruction supplemental interventions.