[en] To investigate the role of PET in predicting the prognosis of resected stage IA non-small cell lung cancer (NSCLC) and planning individualized therapeutic strategies. We retrospectively reviewed the data of patients who underwent surgical resection for lung cancer between January 2004 and December 2014. The clinical data, imaging characteristics of nodules, surgical approaches, and outcomes were analyzed. We evaluated 998 cases; 637 patients with pathological stage I disease were categorized as follows: stage IA1 (251 cases), stage IA2 (250 cases), and stage IA3 (136 cases). The mean follow-up period was 109 months. Significant differences were observed in sex, tumor differentiation, epidermal growth factor receptor mutation, smoking habits, lymphovascular space invasion, tumor size, maximum standard uptake value (SUVmax), and carcinoembryonic antigen level among the groups. Multivariable Cox regression revealed that ground-glass opacity ratio (hazard ratio (HR) = 0.001) and tumor SUVmax independently predicted the postoperative risk of relapse for stage IA3 NSCLC. The HR for SUVmax > 4 was 8.986 (p < 0.001). The 5-year overall survival (OS) rates were 87.2%, 92.9%, and 82.7%, and the 5-year disease-free survival (DFS) rates were 93.2%, 84.2%, and 70.51% for stage IA1, IA2, and IA3 NSCLC, respectively (both p < 0.001). OS and DFS rates were poor in stage IA3 NSCLC patients with an SUVmax uptake > 4 (OS, 71.0% and 92.2%; DFS, 50.2% and 87.3%, for SUVmax > 4 and $\le <!--\; ???\; -->$ 4, respectively; both p = 0.001). SUVmax was a prognostic factor for resected stage IA NSCLC. Postoperative treatment may be considered for IA3 NSCLC with SUVmax > 4.