Third through fifth grade is where students begin to develop their STEM identities to connect their education and the real world. Prior studies of integrated STEM teaching at the middle school level improved students' mathematics self-efficacy and perceived mathematics usefulness. The purpose of this study was to understand how an integrated teaching model (e.g., science, engineering, and mathematics) influenced 5th-grade students' perceptions of their mathematics and engineering abilities. We sought to answer the following research question using a sequential mixed methods research design: how do 5th-grade students' mathematics and engineering self-efficacy and instrumentality for abstract mathematics concepts change because of an integrated teaching experience? We utilized self-efficacy  and perceived usefulness  frameworks to explore how students develop their perceived abilities and usefulness of mathematics through integrated instruction. Seventeen students from a Title-I elementary school worked in teams to design solutions that could provide residents access to clean water. During the integrated unit, students took a pre-, mid-, and post-survey. The items on the survey came from four pre-existing surveys: (1) The Patterns of Adaptive Learning Survey , (2) Mathematical Attitude Assessment , (3) Engineering Skills Self-Efficacy Scale , and (4) Intersectionality of Non-normative Identities in the Cultures of Engineering Survey . The post-interviews captured students' perceptions of their motivations regarding engineering and mathematics. The quantitative and qualitative data created a holistic understanding of how students' perceptions of their abilities shifted throughout the integrated unit. Quantitative data indicated a decrease in self-efficacy but an improvement in perceived mathematics usefulness from mid-unit to post-unit. Qualitative data indicated an increase in students' confidence to do difficult math at the end of the unit. Together results indicated that integrated teaching approaches could foster positive shifts in students' perceived STEM abilities. This kind of instruction could allow students to use and build a broader range of perceived STEM abilities to solve a problem.