A series of studies have assessed the clinicopathological features and prognostic impact of STAS in non-small cell lung cancer (NSCLC) bringing conflicting findings so far. We performed a systematic review and meta-analysis to synthesize the available evidence regarding to the prognostic value of STAS in NSCLCs. Studies were identified by searching databases including PubMed, EMBASE, Web of Science and Cochrane Library up to August 2018 without language restrictions. Results of these searches were filtered according to a set of eligibility criteria and analyzed in line with PRISMA Guidelines. A total of 3754 patients from 14 studies were selected for the present study. The pooled results suggested that presence of STAS was associated with worse RFS (HR=1.975, 95% CI: 1.691-2.307, P <0.001) and OS (HR=1.75, 95% CI: 1.375-2.227, P <0.001) in NSCLCs. Subgroup analysis by histology types indicated the presence of STAS was significantly associated with inferior RFS in resected lung adenocarcinoma (n=7, HR=2.288, 95% CI: 1.843-2.840, I2 = 7.80%), lung squamous cell carcinoma (n=3, HR=1.622, 95% CI: 1.279-2.056, I2 =0%) and lung pleomorphic carcinoma (n=1, HR=4.76, 95% CI: 1.168-19.398). Additionally, a number of clinicopathological characteristics indicating STAS in NSCLCs are summarized. Our study indicates that tumor STAS was a potentially significant prognostic predictor for surgical patients with NSCLCs. The prognostic impact of STAS present in the resection margin remains undetermined. Further large-scale prospective studies are warranted to confirm the prognostic significance of STAS in patients with NSCLCs. Copyright ? 2019. Published by Elsevier Inc.