ObjectivesWe analyze whether the student selection process at Hannover Medical School (MHH), which combined a semi-structured interview with school leaving grades, can predict performances in an Objective Structured Clinical Examination (OSCE). We also check whether there are differences between assessments of clinical knowledge, practical skills, and communication abilities. MethodsWe use data from 525 medical students who were admitted after a successful selection process and who completed the OSCE in the years 2015–2019. We employ multivariate regressions and a mediation analysis approach to learn whether study success after admission and prior to the OSCE mediates the outcome of the latter. ResultsA better performance in the MHH’s selection interview is unrelated to success in the OSCE. However, there is a small but significant influence of school grades on OSCE results in each part except for the assessment of communication skills. The impact of the school grade is partially mediated by performances in written and oral exams preceding the OSCE. DiscussionSchool grades matter for the OSCE outcome, albeit to different degrees for more learning-based vs. practical parts of the examination. The interview at MHH was purely informative and unrelated to study success, also in the assessment of communication skills. Better structured interview tools may yield better results. ConclusionStudents’ cognitive abilities predict study success in an undergraduate OSCE. Performances in a semi-structured selection interview have no impact, not even the assessment of communication skills.