Background: Resuscitative endovascular balloon occlusion of the aorta (REBOA) is a potentially life-saving but high-risk emergency procedure in patients with haemorrhagic shock. Lack of physicians with competence in the procedure is a barrier to implementation of REBOA. It is currently unclear how training and assessment of competence should be done. Objectives: To report and evaluate research in training and assessment of competence in REBOA and femoral arterial access with the aim to investigate the effect of simulation-based training in the procedure and to provide suggestions for the future design of training programs and assessment tools. Methods: Following PRISMA guidelines, PubMed, Embase, and Cochrane Library databases were searched for studies on training or assessment of competence in REBOA and femoral arterial access. Bias assessment was done using the Medical Education Research Study Quality Instrument. Evidence level was assessed using GRADE. Results: Sixteen studies were included, six of them published as abstracts. Full-text studies included 189 trainees ranging in experience level from military medics to surgical specialists. Outcome measures were heterogenous; the most used were rater checklists, knowledge testing, and procedure time. All studies confirmed an effect of training of REBOA on procedural competence in a simulation setting but had a high degree of bias. No study developed or used an assessment tool supported by validity evidence and no study investigated mid and long-term outcomes. Conclusion: Simulation-based training of REBOA improves skills, however, the evidence level is very low and data cannot answer important questions on effect size, skill transfer and retention, and optimal course design. To advance research and training programmes, an assessment tool supported by validity evidence with broad applicability is needed.