To investigate the effectiveness, safety, and prognosis of neoadjuvant chemoradiotherapy (nCRT) for Siewert type II and III adenocarcinomas of the esophagogastric junction (AEG). This study is a prospective randomized controlled clinical study (NCT01962246). AEG patients who were treated at the Third Department of Surgery of the Fourth Hospital of Hebei Medical University from February 2012 to June 2016 were included. All of the enrolled patients were diagnosed with type II or III locally advanced AEG gastric cancer (T2-4N0-3M0 or T1N1-3M0) by gastroscopy and CT before operation; the longitudinal axis of the lesion was ≤ 8 cm; no anti-tumor treatment was previously given and no contraindications of chemotherapy and surgery were found. Case exclusion criteria: serious diseases accompanied by liver and kidney, cardiovascular system and other vital organs; allergy to capecitabine or oxaliplatin drugs or excipients; receiving any form of chemotherapy or other research drugs; pregnant or lactating women; patients with diseases resulting in difficulty to take capecitabine or with concurrent tumors. Based on sample size estimation, a total of 150 AEG patients were enrolled. Using the random number table method, the enrolled patients were divided into the nCRT group and the direct operation group with 75 cases in each group. The nCRT group received XELOX chemotherapy (capecitabine+ oxaliplatin) before surgery and concurrent radiotherapy (45 Gy, 25 times, 1.8 Gy/d, 5 times/week). Clinical efficacy of the nCRT group was evaluated by the solid tumor efficacy evaluation standard (RECIST1.1) and the tumor volume reduction rate was measured on CT. After completing the preoperative examination in the direct operation group, and 8-10 weeks after the end of nCRT in the nCRT group, surgery was performed. Laparoscopic exploration was initially performed. According to the Japanese "Regulations for the Treatment of Gastric Cancer", a transabdominal radical total gastrectomy combined with perigastric lymph node dissection was performed. The primary outcome was the 3-year overall survival (OS) and disease-free survival rate (DFS); the secondary outcomes were R0 resection rate, the toxicity of chemotherapy, and surgical complications. The follow-up ended on December 31, 2019. The postoperative recurrence, metastasis and survival time of the two groups were collected. After excluding patients with incomplete clinical data, patients or family members requesting to withdraw informed consent, and those failing to follow the treatment plan, 63 cases in the nCRT group and 69 cases in the direct operation group were finally enrolled in the study. There were no statistically significant differences in baseline characteristics of the two groups (all >0.05). Sixty-three patients in the nCRT group were evaluated by RECIST1.1 after treatment, the image based effective rate was 42.9% (27/63), and the stable disease rate was 98.4% (62/63); the tumor volume before and after nCRT measured on CT was (58.8±24.4) cm(3) and (46.6±25.7) cm(3), respectively, the effective rate of tumor volume reduction measured by CT was 47.6% (30/63). Incidences of neutrophilopenia [65.1% (41/63) vs. 40.6% (28/69), χ(2)=7.923, =0.005], nausea [81.0% (51/63) vs. 56.5% (39/69), χ(2)=9.060, =0.003] and fatigue [74.6% (47/63) vs. 42.0% (29/69), χ(2)=14.306, =0.001] in the nCRT group were significantly higher than those in the direct surgery group. Radiation gastritis/esophagitis and radiation pneumonia were unique adverse reactions in the nCRT group, with incidences of 52.4% (33/63) and 15.9%(10/63), respectively. The classification of tumor regression of 63 patients in nCRT group presented as 11 cases of grade 0 (17.5%), 20 cases of grade 1 (31.7%), 28 cases of grade 2 (44.4%), and 5 cases of grade 3 (7.9%). Eleven (17.5%) patients achieved pathologic complete response. Sixty-one (96.8%) patients in the nCRT group underwent R0 resection, which was higher than 87.0% (60/69) in the direct surgery group (χ(2)=4.199, =0.040). The mean number of harvested lymph nodes in the specimens in the nCRT group and the direct operation group was 27.6±12.4 and 26.8±14.6, respectively, and the difference was not statistically significant (-0.015, =0.976). The pathological lymph node metastasis rate and lymph node ratio in the two groups were 44.4% (28/63) vs. 76.8% (53/69), and 4.0% (70/1 739) vs. 21.9% (404/1 847), respectively with statistically significant differences (χ(2)=14.552, <0.001, and χ(2)=248.736, <0.001, respectively). During a median follow-up of 52 (27-77) months, the 3-year DFS rate in the nCRT group and the direct surgery group was 52.4% and 39.1% (=0.049), and the 3-year OS rate was 63.4% and 52.2% (=0.019), respectively. According to whether the tumor volume reduction rate measured by CT was ≥ 12.5%, 63 patients in the nCRT group were divided into the effective group (=30) and the ineffective group (=33). The 3-year DFS rate of these two subgracps was 56.6% and 45.5%, respectively without significant difference (=0.098). The 3-year OS rate was 73.3% and 51.5%,respectively with significant difference (=0.038). The 3-year DFS rate of patients with the tumor regression grades 0, 1, 2 and 3 was 81.8%, 70.0%, 44.4%, and 20.0%, repectively (=0.024); the 3-year OS rate was 81.8%, 75.0%, 48.1% and 40.0%, repectively (=0.048). nCRT improves treatment efficacy of Siewert type II and III AEG patients, and the long-term prognosis is good.