Objectives: Urolithiasis is one of the most common presentations in urological practice and it is becoming increasingly important to provide structured, simulation-based training using validated training models. This systematic review aims to identify current simulation-based training models and to evaluate their validity and effectiveness. Methods: Following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, a comprehensive search was performed on the Medline and EMBASE databases for English language articles published between 2000 and 2018 describing and/or assessing validity of simulation models for urolithiasis procedures. Studies were qualitatively assessed for validity using the Messick validity framework and models were assigned levels of recommendation using the McGaghie model of translational outcomes. Results: A total of 98 studies were included in this study assessing 51 models, with 28 studies concerning models for urethrocystoscopy, 46 studies for ureterorenoscopy, and 39 studies for percutaneous access and/or nephrolithotomy. Only four models demonstrated a level of recommendation of 4. The most validated models were the URO/PERC-Mentor (Simbionix, Lod, Israel) with multiple studies for each across various procedural skills. Conclusion: There is a wide spectrum of simulation-based models currently available for urolithiasis procedures, mostly with limited validity evidence from small studies. Further research is required with higher levels of evidence including randomized controlled trials. In addition, long-term transfer of skills to the operating room should be assessed to establish whether there is genuine skill development and retention using simulation models and whether this helps to reduce surgical complications.