Microscopic vascular invasion (MVI) in patients with non-small cell lung cancer (NSCLC) has been reported to be a strong predictor of poor outcomes but it has not been a descriptor of the TNM classification. The purposes of this study were to determine whether the presence of MVI is related to a predictor of poor outcomes and to explore the degree of MVI according to tumor size. A total of 1,884 patients with stage pT1-4N0-2 NSCLC who underwent complete resection comprised the study sample. Overall survival (OS) and recurrence-free proportion were estimated using the Kaplan-Meier method. The Cox proportional hazards model was used to assess independent predictors of poor outcomes. Of 1,884 patients, 1,097 (58.2%) had MVI. Multivariate analysis showed MVI was a significant independent predictor of unfavorable OS (hazard ratio, 1.666; P < .001) and recurrence (hazard ratio, 2.268; P < .001). The frequency of MVI varied according to tumor size, and in each cohort of tumor sizes ≤ 2 cm, > 2 to 3 cm, and > 3 to 5 cm, there were significant differences in survival outcome by MVI status. The proportions of patients with a 5-year recurrence-free period with tumor sizes ≤ 2 cm, > 2 to 3 cm, and > 3 to 5 cm between MVI (+) and MVI (-) were 93.0% and 72.5% (P < .001), 90.8% and 63.3% (P < .001), and 86.4% and 59.9% (P < .001), respectively. This study demonstrated that MVI was a strong predictor of poor outcomes and that the effect is more prominent in patients with tumor sizes ≤ 5 cm. Further analysis of survival and MVI should be collected for future revision of the TNM system.