Background: Albumin-bound paclitaxel has been proven to be an active agent in advanced gastric cancer, and was approved by PMDA Japan for use as a second-line treatment of advanced gastric cancer. The aim of this study was to evaluated the efficacy and safety of AS and SOX in the first-line treatment of advanced gastric cancer. Methods: Patients diagnosed with unresectable locally advanced, recurrent or metastatic human epidermal growth factor receptor type 2 (HER2)-negative gastric or gastroesophageal junction (GEJ) adenocarcinoma were randomized (1:1) to either AS group (albumin-bound paclitaxel 260mg/m² d1 or 130mg/m² d1, 8; S-1 40 mg [BSA < 1.25 m²], 50 mg [1.25 ≤ BSA < 1.50 m²] and 60 mg [BSA ≥1.50 m²] b.i.d. d1-14, every 3wks) or SOX group (oxaliplatin 130mg/m² d1; S-1 40 mg [BSA < 1.25 m²], 50 mg [1.25 ≤ BSA < 1.50 m²] and 60 mg [BSA ≥1.50 m²] b.i.d. d1-14, every 3wks). Primary endpoint was progression-free survival (PFS), secondary endpoints included overall survival (OS), objective response rate (ORR) and safety. Results: Between March 2019 and March 2021, 96 patients were enrolled and 85 patients (AS group, n = 43; SOX group, n = 42) who received at least one dose of study drug and had at least one postbaseline efficacy evaluation were included in the final analysis. 34 (79.0%) patients in the AS group and 35 (83.3%) in the SOX group had discontinued treatment. Of these 85 patients, 78.8% had two or more organs involved, 61.2% had peritoneal dissemination, and more than 1/4 had massive ascites. Patient demographics and disease characteristics were generally balanced between arms. With median follow-up 15.4 months (95% CI, 8.8-21.9), the median PFS of AS group vs. SOX group was 9.2 vs.5.1 months (HR = 0.58 [95% CI 0.36-0.96], p = 0.034); and the median OS was 14.4 vs. 15.4 months (HR = 0.73 [95%CI 0.40-1.34], p = 0.375). Tumor response was assessable in 33 patients according to RECIST V1.1 in both cohorts, the ORR and DCR were similar in AS vs. SOX arm (ORR: 54.5% vs. 51.5%, P = 0.805; DCR: 78.8% vs. 78.8%, P = 1.000). The main grade 3 or worse treatment-related adverse events were neutropenia (30.2% vs 23.8%), Leucopenia (11.6% vs 11.9%) and peripheral sensory neuropathy (4.7% vs 7.1%) in AS and SOX group. The study was planned to be adjusted, due to the widespread use of immunotherapy in first-line treatment of gastric cancer. Conclusions: The results of this study indicated that in previously untreated patients with unresectable locally advanced, recurrent or metastatic HER2-negative gastric or GEJ adenocarcinoma, AS achieved better PFS than SOX, with an inadequate power. The toxicities were controllable and consistent with previously reported. Clinical trial information: NCT03801668.