The Data Puzzles Research Practice Partnership supports middle school science teachers in building confidence to teach data sensemaking using real-world climate data. Through an ongoing professional learning community (PLC) of 19 teachers, the project aims to increase students’ ability to actively make sense of data as they investigate climate phenomena. This work brings together three key areas: climate science education, data sensemaking, and the Data Puzzles instructional framework. At the center of all three is a shared practice—challenging students to explore, analyze, interpret, and reason with authentic datasets to better understand real-world climate issues. The project uses a design-based research approach, focusing on how teachers implement Data Puzzles materials in their classrooms. As part of this work, we collected survey data to understand changes in teachers’ confidence and instructional practices related to both data use and climate science teaching. Our professional learning design emphasizes: (1) using high-quality, standards-aligned instructional materials as a foundation for teacher learning; (2) engaging teachers and students in meaningful work with data; (3) connecting datasets to real-world phenomena; and (4) providing flexible, open-ended opportunities for students to decide how to analyze and represent data. Results show significant increases in teacher confidence across all three areas. Teachers reported higher confidence in data sensemaking, climate science instruction, and use of the Data Puzzles instructional framework after participating in the PLC. The largest gains were seen in familiarity with the Data Puzzles instructional framework. These findings suggest that sustained, practice-based professional learning—grounded in high-quality instructional materials and focused on authentic data use—can support teachers in integrating data science and climate education in meaningful ways.