Female mentoring success was investigated as an undergraduate intervention utilizing career development practices to reduce dysfunctional career thinking and STEM major retention in first year freshmen females within a living-learning community. Repeated measures MANOVAs and canonical correlations in the causal comparative research design evaluated mentoring’s influence on first year females. Male voluntary participants (n = 126) formulated the comparison group, and female voluntary participants (n = 75) filled the treatment group. Repeated measure multivariate analyses of variances compared differences between the interaction of mentoring and gender over time on dysfunctional career thinking using two assessments: Career Thoughts Inventory (CTI) and Career Decision Scale (CDS) and their five subscales (decision-making confusion, commitment anxiety, external conflict, certainty and indecision). Canonical correlations analyzed the effect participation rates had on student change scores on the CTI and CDS, indicating mentoring intervention effects on reducing dysfunctional career thinking and decidedness. Conclusions included: (a) females had higher levels of dysfunctional career thinking than males; (b) overtime both groups decreased dysfunctional thoughts, and solidifying their STEM career choices; (c) females had reduced levels of career decidedness compared to males; (d) both groups increased certainty overtime, solidifying their STEM career choice, and (e) when the STEM career choice was made, female certainty was more solidified than males. The study adds to the career development research within STEM at the undergraduate level providing colleges and universities with a structured first year female mentoring program in STEM. The iv GEMS model may be ideal for colleges and universities utilizing living-learning communities to increase underrepresented female retention and those without STEM career planning courses.