This full, research category study examines how out-of-school experiences in Grades 9-12 predict first-generation college students' engineering possible selves and certainty of career path. The data for this study came from a large-scale survey on outreach programs which was distributed in first-semester English courses to capture an array of responses from students interested in STEM and non-STEM careers. We used structural equation modeling to examine a set of hypotheses: 1) out-of-school experiences would be mediated by interest and recognition in physics and STEM and no direct effect will be found for out-of-school experiences on physics and STEM identities, 2) these identities subsequently predict engineering possible selves, and 3) engineering possible selves will predict certainty of career path. The results of our structural equation modeling analysis supported our hypotheses, out-of-school experiences alone are not enough to develop an identity as a physics person or STEM person, rather they need to be mediated through recognition by others and an underlying interest. A physics identity and a broad STEM identity were found to significantly predict students engineering possible selves. Engineering possible selves were a significant predictor of first-generation college students' certainty of career path. Future possible selves for first-generation college students have important implications for academic development, integration into their community of practice, retention, and the formation of a future professional identity.