Objectives To systematically review the efficacy and safety of non-systemic lymph dissection (NSMLD) vs. systemic lymph dissection (SMLD) for early stage non-small cell lung cancer (NSCLC). Methods PubMed, EMbase, Web of Science and The Cochrane Library databases were searched online to collect randomized controlled trials (RCTs) and non-randomized controlled studies (NRCTs) of NSMLD vs. SMLD for NSCLC patients from inception to October, 2016. Two reviewers independently screened literature, extracted data and assessed the risk of bias of included studies. Meta-analysis was then performed using RevMan 5.3 software. Results A total of 16 studies (4 RCTs and 12 NRCTs) involving 4 718 patients were included. The results of meta-analysis showed that: Compared with the SMLD group, the NSMLD group had higher mortality (HR=1.23, 95%CI 1.11 to 1.37, P<0.000 1). There were no significant differences in disease-free survival, local recurrence rate, distant metastasis rate, and safety between two groups. In addition, the NSMLD group had shorter operation time, and lower drainage and blood loss. Subgroup analysis was performed according to operation methods. The results showed that: NSMLD group by lymph node sampling (LN-S) had higher mortality than SMLD group (HR=1.43, 95%CI 1.17 to 1.75,P=0.004), NSMLD group by lobe-specific lymph node dissection (L-SLD) did not have higher mortality. Conclusions Current evidence shows that: compared with SMLD, NSMLD by L-SLD do not have higher mortality in early stage NSCLC patients, while NSMLD by LN-S have higher mortality. Due to limited quality and quantity of the included studies, more high quality studies are required to verify above conclusion.