Background: Patients with locally advanced gastric cancer (LAGC), are known to have a high risk of recurrence even after surgery. This study was designed to delineate whether neoadjuvant concurrent chemoradiotherapy (CCRT) using the S-1/Cisplatin (SP) regimen can effectively downstage the T stage in patients with clinically T4 stage compared with surgery alone (OP). Methods: This is a randomized phase II open label trial and gastric cancer patients with initial clinical T4 were enrolled. Patients were randomized into neoadjuvant CCRT or OP. For the CCRT group, S-1 (20 mg/m2 twice daily, 2 weeks on and 1 week off) and cisplatin (30 mg/m²/day, first day of week 1, 2, 4, and 5) were administered concurrent with radiotherapy (day 1-5 for 5 weeks, total 45 Gy/25 fractions,). Surgery was performed within 3 to 6 weeks after the completion of CCRT. For the OP group, curative surgery was performed within 3 weeks after randomization. Post-operative adjuvant chemotherapy was chosen as S-1 monotherapy or combination of Capecitabine, and Oxaliplatin according to the physician’s choice. The primary objective of this study was to compare the downstaging rates between the groups (65% and 40% for the CCRT and OP group). Using a type I error rate of 0.10 and a power of 0.80, the planned sample size was 51 patients per group but the study was closed prematurely. Results: Between November 2016 and August 2021, 49 patients with T4 were enrolled. Of those, 26 (CCRT) and 23 (OP) patients were randomized but diagnostic laparoscopy for preoperative screening revealed peritoneal seeding in some patients. Finally, 23 (CCRT) and 17 (OP) patients were evaluable and clinicopathologic characteristics were well balanced between the two groups in per protocol analysis. For CCRT, 19 out of 23 (82.61%) patients satisfied downstage, while among 8 out of 17 (47.06%) patients in OP group satisfied downstage (P=0.038). The median disease-free survival (DFS) was 27.91 months in CCRT group, and 45.96 months in OP group (P=0.985). The median overall survival (OS) was not reached in CCRT group, and 51.2 months in OR group (P=0.335). There were no statistically significant differences between neoadjuvant CCRT and surgery group in the aspect of both hematologic and non-hematologic adverse events. Conclusions: In this study, neoadjuvant SP based CCRT significantly increased downstaging in T stage compared than OP alone. The prespecified genomic analysis with transcriptomic sequencing will be available at the meeting. Clinical trial information: NCT03814759.