Background: Numerous strategies have been applied to combat flat or declining social responsibility (SR) attitudes of undergraduate students in STEM fields. This paper presents the results of a five-year, mixed methods longitudinal study tracking a cohort of undergraduate students at the Georgia Institute of Technology. Drawing on the Professional Social Responsibility Development Model and the Generalized Professional Responsibility Assessment (GPRA), we analyze results from surveys (n = 124 students) and interviews (n = 19 students), each conducted twice at different time points during students' undergraduate education. Our focus is on the influence of different kinds of SR-related activities and experiences on student SR development over time, measured both quantitatively and qualitatively along several dimensions. Based on regression analyses controlling for pre-college attitudes, subgroup and path analyses, and triangulation with qualitative interview data, we identify key influences and inhibitors shaping SR development among STEM undergraduates. Results: Our results reinforce prior findings that student social awareness and professional development remain largely flat over time, and we observe a statistically significant decline in professional connectedness by graduation. We additionally observe a growing emphasis on salary as compared to helping others as a motivating factor in career decisions. Students report rare and unpersuasive efforts as part of in-major and STEM education to build SR, and other formal educational efforts like out-of-major SR activities or activities associated with institutional events have a modest impact as well. In contrast, the most common and positive influences occur outside of formal education. Peer interactions and personal exploration appear to be the most positive forces affecting SR development. Overall, while some students report an increase in basic social awareness, they also tend to report decreases in self-efficacy and few changes to their career orientations related to SR. Conclusions: Despite measurement challenges that occur outside of simpler pre-post-analyses or limited-term interventions, we encourage continued focus on holistic and longitudinal analyses. These may be needed to account for the complex interrelationships between different SR experiences and their impacts on personal and professional SR development. Overall, greater efforts are needed to measure, understand, and improve SR development for undergraduate students. We suggest that it may be important to both enhance the quality of disciplinary SR education in STEM and foster benefits of strategies like peer-based learning.