Background: Currently, NCCN guidelines recommend PD-1 inhibitors for the first-line therapy in gastric cancer patients (pts). However, the role of PD-1 inhibitor treatment in the perioperative setting remains unclear. This phase 2 study aimed to evaluate toripalimab plus FOLFIRINOX or SOX as perioperative treatment regimens in pts with locally advanced resectable G/GEJ. Methods: Pts were 1:1 ratio randomized to A or B groups. Pts in A group received four cycles of toripalimab (240mg, D1, q2w) and FOLFIRINOX regimen (oxaliplatin, 65 mg/m², d1; irinotecan, 150 mg/m², d1; leucovorin, 200 mg/m², d1; and fluorouracil, 2400 mg/m², d1-2, q2w) prior to surgery. Then D2 radical gastrectomy was performed, and another 4 cycles of toripalimab and FOLFIRINOX regimen were given within 2 months after surgery. In B group, chemo was changed to SOX regimen (oxaliplatin, 130 mg/m², d1; S-1, 40-60 mg/m², d1-14, q3w) and toripalimab (240mg, D1) was administered every 3 weeks. The rest treatments were the same as A group. The primary endpoint was tumor regression grade (TRG, TRG 0-1). Key secondary endpoints included pathologic complete response (pCR) rate, 3-year disease-free survival (DFS), 5-year overall survival (OS) and safty. Results: Up to Jan 28, 2023, 54 eligible pts were enrolled (A 21, B 33) with baseline characteristics as follows: The median (range) age was 67 (23-71) years; 6/54 (11.1%) pts had signet ring cell carcinoma component; 40/54 (74.1%) pts had cT4 tumors and 18/54 (33.3%) had lymph node positive disease. 32/54 (59.3%) pts underwent successful resection. R0 resection rate was 100%. The TRG 0-1 rate was higher but not significant in the B group than in the A group (31.58% vs. 23.08%, P = 0.703). 4 pts (A 2, 15.4%; B 2, 10.5%) had a pCR. 23 (A 8, B 15, 71.9%) pts had tumor downstaged after surgery. Treatment-related adverse events (TRAEs) of any grade occurred in 25/54 (46.3%) pts. Grade ≥3 TRAEs occurred in 10/54 (18.5%) pts: 7 with grade 3-4 neutropenia, 2 with grade 3 thrombocytopenia and 1 with grade 3 myelosuppression. No new safety signals were observed and no TRAEs leading to death in either group. Conclusions: Toripalimab plus FOLFIRINOX or SOX regimens is tolerable and effective for pts with locally advanced resectable G/GEJ in the perioperative treatment setting. There were no significant differences in efficacy and safety between the two groups. Whether intensive or simplified treatment is more suitable for neoadjuvant therapy needs to be further explored. The low pCR rate in both groups may be due to the small sample size and the pts with more advanced primary tumors. This study is ongoing and more data will be released. Clinical trial information: NCT04908566.