The study investigated the value of preoperative magnetic resonance cholangiopancreatography (MRCP) in predicting the success of removal of common gallstones (CBDS) by endoscopic retrograde cholangiopancreatography (ERCP). It included 432 patients, of which 175 were men and 257 were women. Stones were successfully removed in 395 patients (success rate 91.44%). There were significant differences between the successful and unsuccessful groups in stone diameter (p < 0.001), the widest diameter of the common bile duct (CBD, p < 0.001), and the ratio of the widest diameter of the CBD/stone diameter (p < 0.001), but not in the number of patients with one stone (p = 0.174). Binary logistic regression showed that stone diameter > 15 mm was an independent risk factor for ERCP failure (OR = 11.229, 95% CI: 1.576–80.033, p = 0.016). ROC curve analysis for this cutoff achieved an AUC of 0.94, a sensitivity of 100%, and a specificity of 75%. Preoperative MRCP thus significantly helps to evaluate the success of ERCP in patients with CBDS.