<p>Immediate basic life support (BLS) performed by medical staff in general wards enhances patient outcomes after in-hospital cardiac arrest. To ensure the quality of the BLS, high-frequency trainings are necessary. A training approach that is independent of personnel resources and can still effectively train staff is virtual reality led BLS training. Therefore, this non-inferiority study investigated whether VR–BLS training is as effective in training medical staff as traditional mannequin-based training. A randomized controlled trial was conducted in 2023 in Hamburg, Germany. Medical staff working in general wards were randomly assigned to either a control group or an intervention group for BLS training. The control group underwent traditional mannequin-based training, while the intervention group received VR-based training. Both groups participated in an objective Structured Clinical Exam (OSCE) following their training. The primary endpoint was the mean score in a BLS assessment (with a total of 220 penalty points) during the OSCE, and the secondary endpoint was the subjective learning gain. The BLS assessment and subjective learning gain of a total of 106 medical staff members were analyzed. Participants in the classical (mannequin-based) training received a total of 3.302 penalty points less (95%-CI [− 3.581; 10.122]), t (209) = .954, <i>p</i> = .341 in the OSCE than the intervention group. Although statistical non-inferiority could not be confirmed, the differences in performance were clinically negligible. Integrating all results of this study, VR based BLS training is an effective alternative to the classical mannequin-based approach, facilitating individual high-frequent BLS trainings.</p>