Objective: The present study investigated the efficacy of the didactic approaches of video modeling (VM, best-practice examples), video reflection (VR, problem-based approach), and the combination of both (VMR) in fostering medical communication competence in a video-based digital learning environment. Methods: N = 126 third-year medical students who participated in the pre-post study were assigned to either the intervention groups (VM, VR, and VMR) or the wait-list control group. The efficacy of the three approaches was assessed by means of a situational judgment test (SJT) of medical communication competence. To investigate the differences between the wait-list control and the intervention groups (H₁), between the single-mode and combined conditions (H₂), and between VM and VR (H₃), we applied planned contrast analyses. Results: The planned contrasts showed that the VR condition significantly improved learning outcomes in comparison to the VM condition (H₃). The decreased mean scores of the VM condition offset the increased mean scores of VR, and thus no significant differences could be found in H₁ and H₂. Conclusions: Our study provides promising evidence that VR fosters medical communication competence more effectively than VM. Practical implications: Medical students’ learning in video-based digital environments can be facilitated by the use of interactive VR.