Background: In the Montreal classification, L4 Crohn's disease (CD) is defined as an ileal disease, including L4-esophagogastric duodenum (EGD), L4-jejunum, and L4-proximal ileal involvement. According to the previous studies, the prognosis of L4 disease was worse than that of non-L4 disease. Among L4 diseases, the phenotypes of L4-jejunum and L4-proximal ileum indicated that the risk of abdominal surgery was higher. However, the prognosis of L4-esophagogastroduodenal remains largely elusive. Therefore, we aim to investigate whether the prognosis differs between CD patients with and without esophagogastroduodenal involvement. Methods: In this study, patients with L4-EGD phenotype (n = 74) who underwent gastroscopy, ileocolonoscopy, biopsies, and CTE from 2018 to 2020 were compared with L4 non-EGD controls (n = 148) who were randomly selected at a ratio of 1:2 in the same period. Demographic information inclusive of disease conduct and location, important points of the surgery, and hospitalization have been collected. The distinction between L4-EGD patients and non-L4-EGD patients was identified by way of multivariable logistic regression analysis. The Kaplan-Meier technique was used to consider the possibility of abdominal surgical operation and complications, observed by means of Cox percentage hazard fashions to decide if L4 EGD independently estimated the endpoints inclusive of the abdominal surgery and the occurrences of complications. Results: L4-EGD group (n = 74) had a lower proportion of intestinal fistula than the control group (n = 148) (17.6% versus 34.5%; p = 0.009), and the probabilities of requiring abdominal surgery and multiple abdominal surgeries were also lower (21.6% versus 36.5%; p = 0.025), (6.8% versus 18.9%; p = 0.016), respectively. The frequency of hospitalization was lower in the L4-EGD group than in the control group (3-7 versus 4-9; p = 0.013). L4-EGD phenotype was found to be an independent protective factor for abdominal surgery and intestinal fistula in the Cox regression model, with HRs of 0.536 (95%CI: 0.305-0.940; p = 0.030) and 0.478 (95%CI: 0.259-0.881; p = 0.018), respectively. Conclusion: Our data suggest that the L4-EGD phenotype may have a better prognosis compared to the Non-L4-EGD phenotype. Our data may advocate a revision of the Montreal classification including separate designations for L4-EGD disease.