BACKGROUND: The management of ground glass opacity (GGO) on computed tomography (CT) remains controversial. Information of the relationship between clinical behavior and pathological invasiveness of GGO is valuable for management. We conducted this retrospective study to establish differences in the pathological invasiveness between GGO with and without changes. METHODS: Among 1762 patients, the following criteria was used: (1) maximum tumor diameter of 3 cm or less, (2) tumor having 50% or more GGO and (3) resection after at least three months of follow up. A change of CT findings was defined as an increase in the diameter or consolidation compared with the initial CT. The relationship between preoperative changes and ratio of invasive adenocarcinoma was investigated. Predictors of GGO growth were also examined. RESULTS: There were 250 patients: pure GGO without changes (G-N group; n = 118), pure GGO with changes (G-C group; n = 35), part-solid GGO without changes (S-N group; n = 78), and part-solid GGO with changes (S-C group; n = 20). The ratio of invasive adenocarcinoma in each group was 0.54, 0.89, 0.8, and 0.90. There was a significant difference between the G-N and G-C group (P < 0.001). However, there was no significant difference between the G-C, S-N and S-C group. Multivariate analysis indicated age was a predictor of preoperative changes (OR = 1.953, P = 0.049). CONCLUSIONS: The pathological results of part-solid GGO with changes were not different from those without changes. Therefore surgery can be deferred until those lesions demonstrate changes. The pathological results of pure GGO with changes were equivalent to those of part-solid GGO. Therefore, even for pure GGO, follow up is necessary especially in elderly patients.