Medical school can be a socially isolating experience, particularly for students underrepresented in medicine. Social isolation and perceptions of not belonging can negatively impact students' academic performance and well-being. Therefore, interventions are needed to support students and these efforts should be appealing, brief, and low-burden. Guided by evidence-based approaches, we developed the Build & Belong intervention for medical students as a brief peer-to-peer approach that consisted of four components. First, M3 and M4 students wrote reflections on belonging in medical school. Second, M3 and M4 students video recorded messages for M1 and M2 students using their written reflections. Third, M1 and M2 students watched and discussed the videos in small groups. Fourth, the M1 and M2 students wrote letters to future students. Our intervention differs from previous student belonging interventions in the peer delivery of messages. The Build & Belong intervention aimed to improve medical students' social belongingness. Using a longitudinal observational study design, the intervention was piloted at a medical school in the Mid-Atlantic United States in 2017-2018. Students completed surveys before and after the intervention. Paired samples tests (t-tests and Wilcoxon) assessed pre- to post-intervention changes in social isolation, social connectedness, and social assurance. Among 63 medical students, with 25.9% from backgrounds underrepresented in medicine, we assessed follow-up outcomes in 38 students. Social isolation scores significantly decreased from baseline (M = 54.8, SD = 7.06) to follow-up (M = 51.3, SD = 6.67; p < .001). Social isolation changes were evident regardless of sex, although males reported a greater reduction (M Δ = −5.32, p < .001) than females (M Δ = −2.79, p = .014). Black/African American students had the largest reduction in social isolation (M Δ = −7.24, p = .010). Social assurance and connectedness scores did not change significantly between baseline and follow-up. Medical students appeared to resonate with messages delivered by more experienced peers (M3s and M4s), particularly messages that normalized feelings of not belonging and strategies to reduce those feelings. The Build & Belong intervention appears to reduce social isolation scores among medical students. This pilot test of the Build & Belong intervention provides initial evidence of the effectiveness of a brief, low-cost intervention. Build & Belong may provide a scalable strategy to reduce medical students' social isolation. Our peer-based approach is distinct from administrator-led strategies; peers were seen as trusted and reliable sources of information about belonging and ways to overcome the challenges experienced during medical school.