BACKGROUND: Increasing the attractiveness of primary care careers is a key step in addressing the growing shortage of primary care physicians. The purpose of this review was to (1) identify interventions aimed at increasing the proportion of undergraduate medical students choosing a primary care specialty, (2) describe the characteristics of these interventions, (3) assess the quality of the studies, and (4) compare the findings to those of a previous literature review within a global context. METHODS: We searched MEDLINE, EMBASE, ERIC, CINAHL, PsycINFO, The Cochrane Library, and Dissertations & Theses A&I for articles published between 1993 and February 20, 2015. We included quantitative and qualitative studies reporting on primary care specialty choice outcomes of interventions in the undergraduate medical curriculum, without geographic restrictions. Data extracted included study characteristics, intervention details, and relevant outcomes. Studies were assessed for quality and strength of findings using a five-point scale. RESULTS: The review included 72 articles reporting on 66 different interventions. Longitudinal programs were the only intervention consistently associated with an increased proportion of students choosing primary care. Successful interventions were characterized by diverse teaching formats, student selection, and good-quality teaching. Study quality had not improved since recommendations were published in 1995. Many studies used cross-sectional designs and non-validated surveys, did not include control groups, and were not based on a theory or conceptual framework. DISCUSSION: Our review supports the value of longitudinal, multifaceted, primary care programs to increase the proportion of students choosing primary care specialties. Isolated modules or clerkships did not appear to be effective. Our results are in line with the conclusions from previous reviews and add an international perspective, but the evidence is limited by the overall low methodological quality of the included studies. Future research should use more rigorous evaluation methods and include long-term outcomes.