OBJECTIVES: The role of segmentectomy in early-stage non-small-cell lung cancer (NSCLC) remains a matter of debate. We performed a meta-analysis to evaluate the oncological outcomes following segmentectomy versus lobectomy for stage I, stage IA only and stage IA <2 cm only. METHODS: We systematically searched the literature for articles reporting on overall survival (OS), cancer-specific survival (CSS) or recurrence-free survival (RFS). The hazard ratios (HRs) were retrieved and pooled using an inverse variance-weighted approach. RESULTS: Twenty-eight studies were included in the analysis. In stage I, segmentectomy was found to be inferior to lobectomy for all 3 outcomes with HR: 1.25 (P = 0.01) for OS, 1.59 (P = 0.02) for CSS and 1.40 (P < 0.001) for RFS. In stage IA, the differences were significant for OS and CSS, though not for RFS with HR: 1.31 (P = 0.04), 1.56 (P = 0.02) and 1.22 (P = 0.11), respectively. In stage IA <2 cm, no significant differences were found between segmentectomy and lobectomy with HR: 1.13 (P = 0.37) for OS, 1.02 (P = 0.95) for CSS and 1.24 (P = 0.11) for RFS. CONCLUSIONS: For stages I and IA, lobectomy showed superior results whereas for tumours <2 cm, our study did not find significant differences in oncological outcomes between both groups. These results suggest that segmentectomy might be a valuable alternative to lobectomy for NSCLC in tumours <2 cm.