Background The current gold standard for gastric cancer (GC) screening is pathology or a barium meal followed by X-ray. This is not applicable to a wide range of screening capabilities due to the lack of operability. This article used a meta-analysis to evaluate the value of pepsinogen (PG) screening for GC. Methods PubMed, EMbase, the Cochrane Library, CNKI, WanFang, VIP, and CBM databases were systematically searched for published studies that used serum PG to diagnose GC. Articles were searched from January 2003 to January 2018. Two reviewers independently screened the literature according to specified inclusion and exclusion criteria. The data were extracted and evaluated, and the quality of the methodologies evaluated using the QUADAS entry. The meta-analysis (MA) was performed using Meta-DiSc 1.4 software. Stata 12.0 software was used to assess publication bias. Results A total of 19 studies were finally included from a total of 169,009 cases. The MA showed a combined sensitivity and specificity of 0.56 (95% CI (0.53–0.59), P < 0.01) and 0.71 (95% CI (0.70-0.71), P < 0.01), respectively. The combined likelihood ratios were +LR = 2.82 (95% CI (2.06–3.86), P < 0.01) and −LR = 0.56 (95% CI (0.45–0.68), P < 0.01). The combined DOR was 5.41 (95% CI (3.64~ 8.06), P < 0.01), and the area under the SROC curve was 0.7468. Conclusions Serum PG provides medium levels of sensitivity and specificity for GC assessment. To be used in a clinical setting, further high-quality research must be performed and verified.