In a randomized-controlled trial, we tested 2 brief interventions designed to mitigate the effects of a "chilly climate" women may experience in engineering, especially in male-dominated fields. Participants were students entering a selective university engineering program. The "social-belonging intervention" aimed to protect students' sense of belonging in engineering by providing a nonthreatening narrative with which to interpret instances of adversity. The "affirmation-training intervention" aimed to help students manage stress that can arise from social marginalization by incorporating diverse aspects of their self-identity in their daily academic lives. As expected, gender differences and intervention effects were concentrated in male-dominated majors (<20% women). In these majors, compared with control conditions, both interventions raised women's school-reported engineering grade-point-average (GPA) over the full academic year, eliminating gender differences. Both also led women to view daily adversities as more manageable and improved women's academic attitudes. However, the 2 interventions had divergent effects on women's social experiences. The social-belonging intervention helped women integrate into engineering, for instance, increasing friendships with male engineers. Affirmation-training helped women develop external resources, deepening their identification with their gender group. The results highlight how social marginalization contributes to gender inequality in quantitative fields and 2 potential remedies.