Objective The purpose of this systematic review and meta-analysis was to examine the effects of simulation-based ultrasound training (SIM-UT) in obstetrics and gynecology compared to non-SIM-UT on trainee learning, clinical performance, patient-relevant outcomes, and cost of training. Methods A systematic search was performed in June 2019 in PubMed, Embase, and Scopus using search terms for the topic and the intervention as well as certain MESH terms. Inclusion criteria were defined in accordance with the PICO question. Studies published in any language involving SIM-UT in obstetrics and gynecology compared to non-SIM-UT or no training were included. The outcomes included effects on health care provider learning and clinical performance, patient-relevant outcomes, and cost of training. Two authors evaluated the study quality with the MERSQI instrument and the Oxford Quality Scoring System. A meta-analysis was planned for the included randomized controlled trials. Results 15 studies were included, and 11 studies were eligible for meta-analysis. SIM-UT was significantly superior to clinical training only and theoretical teaching with standard mean differences (SMD) of 0.84 (0.08-1.61) and 1.20 (0.37-2.04), respectively. However, SIM-UT was not superior to live model training; SMD of 0.65 (-3.25-4.55). Of all studies included in the meta-analysis, 91 % favored SIM-UT over clinical training alone, theoretical teaching, or in some cases live model training. Conclusion In the field of obstetrics and gynecology, SIM-UT in addition to clinical training markedly improves trainee learning, clinical performance, as well as patient-perceived quality of care.