Teacher professional development and in-class mentors were used to support structured inquiry with math manipulatives. Twenty-four primary schools (n = 6,628 students) were randomly assigned to treatment and control groups as an experimental field trial to examine the effectiveness of this instructional approach in a scaled-up application in Belize for the duration of a school year. Implementation fidelity measures were collected permitting evaluation of two separate multilevel models: intention-to-treat and test-of-treatment. Both quantitative and qualitative evidence suggest students within this culture respond well to this relatively simple and inexpensive intervention that departs from traditional, expository math instruction in many developing countries. Policy implications are discussed that supported nationwide rollout of the intervention.