Despite gains in higher educational attainment, women remain underrepresented in STEM disciplines, a disparity that disenfranchises women from higher earnings later in life. Scholarship from across the social sciences has uncovered the gendered meanings associated with academic fields, but little work has attended to how these gendered associations surface – and may be enacted – during the college admissions process. My research takes advantage of 240,000 undergraduate applications to a multi-site US university, and pairs computational-linguistic methods with theoretical insights from sociology of gender and higher education. I am interested in uncovering novel sites of gender (and race and class) bias embedded in undergraduate application materials, and how gender might inform differential narrations of academic ambition in personal statements. My work has broad implications for gender equity in education and work, and for the utility and ethics of algorithmic approaches in future research and practice.