We report on a practitioner implementation of Fusion, a first-grade mathematics intervention. Studies such as this evaluation of a loose implementation under realistic conditions are important to curriculum developers' understanding of how evidence-based programs and tools work under a variety of implementation scenarios. In this quasi-experimental study, first-grade students (n = 253) in 10 schools were assigned to treatment (n = 154) or control (n = 99). Rather than randomly assigning students, schools assigned those students most at risk to treatment and, typically, those less at risk to control. School staff administered pre- and postassessments and led Fusion sessions approximately 30 min per day, 3 days per week. The intervention resulted in a significant positive effect on a researcher-developed first-grade math measure. The implementation of Fusion and feedback from school staff provided insights to guide the curriculum development process.