The "Coding Our Future" intervention aimed to develop and implement a Computer Science (CS) curriculum, and to train teachers to offer computer science courses in elementary and middle schools. Its goals were to enhance students' competency and self-efficacy in computer science and to boost the enrollment of underrepresented students in these courses. This study was a cluster-level randomized controlled trial (RCT). Treatment and control groups were randomly assigned within schools at grade-level. Within 26 San Diego Unified School District (SDUSD) elementary schools, each with 3rd and 4th grade levels, we randomly assigned grade levels to either the treatment or comparison condition. This means that one grade level at each school received both the Science, Technology, Engineering, Art, and Mathematics (STEAM) units and the CS intervention, while the other grade level received STEAM units only. The primary outcomes examined in this RCT were academic achievement in reading and math, as measured by the FastBridge Reading and Math assessments, and students' computational thinking skills. Additional outcomes, not eligible for evaluation by the WWC, included students' interest in computer science and computer science careers. Results of hierarchical linear models demonstrated that participating in the CS intervention did not significantly impact academic achievement in reading or math. However, our implementation fidelity study indicated that, while professional development attendance met fidelity, 60% of teachers who should have implemented curricular units did not. Therefore, Coding our Future units were implemented with fidelity and it is likely that their true effect was not captured in this study. Despite these insignificant findings, students enjoyed the curriculum, asked for more challenges, and had a unique opportunity to learn with their teachers.