Medical students and doctors have to be particularly stress-resilient, as both medical education and practice are considered very stressful. Specific stressors can lead to increased risks of developing, for example, depression, anxiety and burnout. Relaxation techniques have proven to be effective for the prevention of these outcomes in student populations. However, only a very few medical students practice relaxation techniques regularly early on in their studies. Furthermore, it is unclear which students make use of stress-management offers and hence whether vulnerable students are generally reachable. Therefore, the aim of our study was to explore predictors of participating in a voluntary stress management course for sophomore medical students. One cohort of freshmen at a German medical school was surveyed at the end of the freshman year [t1] and at the end of the sophomore year [t2]. In addition to sociodemographic information, we captured perceived study stress, self-rated general health and mental health and dimensions of study-related behaviour and experience as potential predictors of participation at t1. During the sophomore year, we offered the participants a progressive muscle relaxation (PMR) beginners' course. At t2, we registered participation status. We used binary logistic regression analyses in order to assess correlations between potential predictors and participation. About one third of the whole class took part in the course. The main reason for non-participation was "no time". Being female and higher levels of anxiety were the strongest predictors of course participation. Career ambition (the higher, the less likely to participate) and emotional distancing (the higher, the more likely to participate) were further significant predictors. Future interventions should be attractive to both male and female medical students. Ideally, for every hour of stress management teaching, the curriculum should be cut by at least the same amount of time.