Endoscopic retrograde cholangiopancreatography (ERCP) is technically demanding and carries significant risks. It is performed by gastrointestinal and surgical endoscopists. There is no consensus on the minimum number of ERCPs required during training. This study was conducted to analyze the minimum number of clinical ERCPs that a trainee needs to perform to achieve competency. PubMed, Ovid-Embase, and the Cochrane library were searched systematically for prospective and retrospective studies reporting on trainees’ ERCP performance. Mete-analysis was conducted to analyze the success rate of cannulation, other basic techniques, and adverse event rate, using the random-effect model with Review Manager 5.3. Thirteen studies met the inclusion criteria, with 149 trainees performing a total of 18 794 ERCP procedures. The pooled cannulation success rate was 85.7% (95% CI: 78.1%-91.0%) at completion of training. The cannulation success rate was 76.5% (95% CI: 69.2%-82.5%) when the trainees had completed 180 ERCPs, which increased to 81.8% (95% CI: 69.8%-90.6%) after 200 ERCP procedures. Adverse events and post-ERCP pancreatitis rates were 4.7% (95% CI: 2.9%-9.1%) and 2.0% (0.9%-3.9%), respectively. Achieving a cannulation success rate of >90% was considered a quality indicator for ERCP training by most societal guidelines. However, our retrospective analysis indicated that trainees only attained a pooled cannulation success rate of only 81.8% after 200 procedures. Therefore, the minimum number of ERCPs required to achieve competency during training may need to be redefined to meet the basic requirement.