In patients with advanced non-small cell lung cancer (NSCLC), the benefit-to-risk ratio of doublet-targeted agents versus single agent is not clear. A systematic review and quantitative meta-analysis were, therefore, undertaken to evaluate the available evidence from randomized trials. This study aims to evaluate the efficacy and safety of erlotinib versus doublets (erlotinib plus another targeted agent) in advanced NSCLC and, if adequate data are available, to investigate whether or not predefined patient groups benefit more or less from doublet-targeted therapy based on erlotinib. Medline, Embase, and the Cochrane Central Register of Controlled Trials were searched. Randomized controlled clinical trials were conducted in which any erlotinib was compared with doublets based on erlotinib in patients with NSCLC who had failed to respond to any previous chemotherapy regimen. Two review authors independently selected studies for inclusion in the review and extracted data. A systematic review and meta-analysis based on aggregate data extracted from trial publications were carried out to assess the effectiveness of doublets (erlotinib plus another targeted drug) in NSCLC treatment. The efficacy outcomes were objective response rate (ORR), complete response plus partial response; disease control rate (DCR), complete response plus partial response and stable disease; and 1-year overall survival (OS). The adverse effects (AEs) were also considered. This involved identifying eligible randomized controlled trials (RCTs) and extracting aggregate data from the reports of these RCTs. Hazard ratios were calculated from published summary statistics and then combined to give pooled estimates of treatment efficacy. This meta-analysis comprised five studies including 2,100 patients (mean age 63; 1,224 men and 876 women; 118 stage IIIB and 1,180 stage IV; 441 squamous cell cancers, 1,287 adenocarcinomas, and 372 other pathological types). Doublets regimen significantly improved ORR [hazard ratio (HR) 1.49, 1.13-1.98, p < 0.05] and DCR (HR 1.25, 1.12-1.39, p < 0.05) compared with single erlotinib, but 1-year OS was not significantly improved for doublets [HR 1.06; 95 % confidence interval (CI), 0.95-1.18]. All-grade rash, anemia, diarrhea, anorexia, and fatigue were not significantly different between doublet and erlotinib groups (HR 1.25, 0.99-1.58; 0.98, 0.78-1.24; 1.43, 0.97-2.11; 1.18, 0.84-1.65; and 1.23, 0.86-1.77, respectively). The total grade of ≥3 AEs was also not significantly different (HR 1.40, 95 % CI 0.97-2.01). Compared with single-agent erlotinib, doublets (erlotinib plus another targeted agent) significantly improve ORR and DCR, but not OS, and induce no significance of more frequent and serious AEs. The benefit-to-risk ratio of doublets in advanced NSCLC may be more favorable than that of single-agent. The results of this systematic review suggest that patients with advanced NSCLC might benefit from doublet-targeted therapy based on erlotinib compared to erlotinib alone. However, an individual patient data systematic review and meta-analysis are needed to give us a more reliable assessment of the size of benefits and to explore whether doublet therapy may be more or less effective for particular types of patients.