The purposes of this study were (a) to test the efficacy of whole-class equal-sign instruction, (b) to contrast the effects of symbolic and nonsymbolic instructional representation, (c) to determine whether visual representation compensated for language deficits, and (d) to evaluate maintenance of intervention. We randomly assigned 195 children from 21 second-grade classrooms to 3 conditions: symbolic intervention, nonsymbolic intervention, and business-as-usual control. Both active intervention conditions consisted of three 20-minute instructional lessons. The only distinction between the intervention conditions was problem representation (symbolic vs. nonsymbolic); verbal instruction remained consistent. Children in both instructional conditions outperformed children in the control condition on all outcome measures. Language ability moderated the symbolic intervention relative to control and nonsymbolic intervention, suggesting that nonsymbolic representation may compensate for language deficits.