Purpose: To prospectively determine the feasibility of 3.0-T MR to evaluate the signal characteristics of normal esophageal wall and subsequently assess the depth of esophageal wall invasion by esophageal carcinoma. Materials and Methods: Written informed consent was obtained from all patients and the study was reviewed by the institutional review board. Between January 2015 and April 2016, 41 consecutive patients each contained an esophageal carcinoma were studied on a 3.0-T MR scanner. Three independent readers who were blindly to the histopathological findings reviewed the MR images for evaluating the carcinoma invasion. The stage determined by MR images were compared with the histopathological findings, which were served as the standard reference. Spearman correlation coefficient analysis was used to compare the stage at MR imaging with that at histopathological analysis. A P value less than 0.05 was considered to indicate a significant difference. The interobserver agreement were compared with Cohen k. Sensitivity, specificity and accuracy for carcinoma invasion were calculated. Results:High-resolution MR images clearly depicted the normal esophageal wall as consisting of eight layers, which correlated well with the histopathological layers (Figure 1). In 36(87.8%) of 41 lesions (Figure 2), the depth of esophageal wall invasion determined with MR were consistent with the histopathological stage (r=0.958, P<0.001). Sensitivity, specificity and accuracy for mucosa were 66.7%, 97.1%, 92.7%; for submucosa were 80%, 90.3%, 87.8%; for muscularis propria were 91.7%, 96.6%, 95.1%; for adventitia were 100%, 100%, 100%(Table 1, 2). Cohen k values of interobserver agreement was excellent (K=0.832, P<0.001(observer 1 vs observer 2); K=0.900, P<0.001(observer 1 vs observer 3); K=0.866, P<0.01(observer 2 vs observer 3)). Conclusion: High-Resolution ex vivo MR images can be used to evaluate the esophageal carcinoma invasion precisely and provides good diagnostic sensitivity, specificity and accuracy.