Cetuximab presents a potential therapy for gastric or esophagogastric junction adenocarcinoma. We aim to evaluate the predictive value of potential biomarkers of cetuximab efficacy. In this prospective phase 2 trial (NCT00477711), we enrolled untreated 47 patients with un-resectable or metastatic gastric or esophagogastric junction adenocarcinoma from seven sites in China. Patients with histologically confirmed adenocarcinoma were given cisplatin (80 mg/m2, triweekly), capecitabine (2,000 mg/m2, triweekly for 2 weeks), and cetuximab weekly (400 mg/m2 at first infusion and 250 mg/m2 subsequently). Sample size was calculated using Simon's two-stage design. The primary endpoint was the objective response rate (ORR). Secondary endpoints were progression-free survival (PFS), overall survival (OS), toxicity, and predictive biomarkers. The ORR was 53.2%, median PFS 5.2 months, and OS 10.8 months. The most frequent toxicities included neutropenia (25.0%), nausea/vomiting (11.5%), and rash/desquamation (9.6%). Patients with grade 2-4 rash achieved a significantly better ORR, longer PFS, and OS than those with grade 0-1 rash. Seven patients (15.9%) with epidermal growth factor receptor (EGFR) strong expression (3+) showed great tumor shrinkage, longer PFS (7.1 months), and OS (16.6 months). EGFR gene amplification was detected in four patients (8.5%), all of whom responded well. Compared to patients with lower levels of transforming growth factor-alpha (TGF-α), those with high levels showed better response and longer PFS (6.0 vs 2.7 months, p=0.001) and OS (12.9 vs 7.0 months, p=0.001). C+XP was well tolerated and effective for advanced gastric or esophagogastric junction adenocarcinoma as first-line therapy. Severity of skin rash and TGF-α level correlated with efficacy, and EGFR overexpression might predict cetuximab efficacy.