<p>Engineering problem solving imposes early challenges to students’ self-efficacy beliefs, particularly for students underprepared for the mathematics demands of engineering majors. Grounded in an integrative view of STEM education, we examined change in non-calculus-ready students’ (<i>N</i> = 107) engineering self-efficacy across an introduction to engineering reasoning course. Using Bayesian latent growth curve modeling, we modeled interindividual variability in engineering self-efficacy across three time points of students’ first semester of a fundamentals of engineering program. We also tested the long-term associations between students’ pre-course help-seeking motives and engineering interest, change in engineering self-efficacy, and indicators of their social-academic success. Students’ engineering self-efficacy increased across the course, and supported individual differences in change in self-efficacy were observed. Students’ pre-course engineering interest and help-seeking motives positively predicted initial levels of engineering self-efficacy but not their growth in self-efficacy over time. Growth in engineering self-efficacy was positively associated with students’ end-of-course belonging and mathematics self-efficacy and end-of-year mathematics grades and cumulative GPA. We consider implications for early, efficacy-promoting curricular supports in engineering.</p>