BACKGROUND The aim of this study was to evaluate the efficacy of RASSF1A promoter hypermethylation of serum or sputum in diagnosis of non-small cell lung cancer (NSCLC) by pooling open published data. MATERIAL AND METHODS Open-published studies relevant to RASSF1A promoter hypermethylation and NSCLC diagnosis were screened through Medline, EMBASE, the Cochrane Library, Web of Science, Google Scholar, and CBM. Number of cases of true positive (tp), false positive (fp), false negative (fn), and true negative (tn) by RASSF1A gene promoter hypermethylation was extracted from each of the include original studies. The combined diagnostic sensitivity, specificity, and symmetric receiver operating characteristic curve (SROC) were calculated, as was the effect size. RESULTS Twelve studies with 826 NSCLC and 598 controls were included in the present work. The combined sensitivity and specificity were 0.45 (95%CI: 0.41-0.48) (random effects) and 0.99(95%CI: 0.98-1.00) (fixed effects) respectively. The pooled positive likelihood ratio (+LR) and negative likelihood ratio (-LR) were 20.27 (9.64-42.61) and 0.53 (0.42-0.66), respectively, through the random effects model. The combined DOR was 46.63 (95%CI: 17.30-125.65) through the fixed effects model. The AUC of the SROC was 0.9989, calculated through Moses's model for RASSF1A promoter hypermethylation as a biomarker in diagnosis of NSCLC. CONCLUSIONS The low diagnostic sensitivity for RASSF1A gene promoter hypermethylation indicated that it is not suitable for NSCLC screening. However, the high specificity made it effective for NSCLC confirmation diagnosis, which could be used instead of pathological diagnosis.