Surgeons have long been interested in sublobar resection for non-small cell lung cancer (NSCLC). Preoperative prediction of lymph node metastases is essential for selecting the right patient for sublobar resection. Even in stage I NSCLC, lymph node metastases may sometimes be found on pathological examination. Therefore, systematic lymph node dissection or sampling is recommended even for stage-accurate NSCLC with small volume lymph node negativity. However, if we can correctly predict patients with lymph node negative NSCLC, then systemic lymph node dissection is not necessary in these patients; i.e., sublobar resection (wedge resection or segmental lung resection) without systemic lymph node dissection becomes possible. The authors have reported the usefulness of high-resolution CT (HRCT) and PDG-PET/CT for predicting early lung adenocarcinoma. However, the predictors of preoperative lymph node metastasis in squamous lung cancer are not well understood. To discover predictors of lymph node status in stage IA lung squamous carcinoma, Professor Morihito Okada of Hiroshima University, Japan, investigated this issue, but no suitable predictors were found, and the article was recently published in EUR J CARDIO-THORAC. In this study, 100 patients with stage IA squamous lung cancer who underwent lobectomy or segmental lung resection plus systemic lymph node dissection were analyzed. Lymph node status and tumor size on CT, SUVmax values on PDG-PET/CT, and serum tumor markers were analyzed. Lymph node metastasis was found in 12% of the study subjects and pathological tumor invasion, such as lymph node metastasis and/or lymphovascular, vascular or pleural invasion, was found in 54% of the study subjects. Tumor size, SUVmax value, serum carcinoembryonic antigen, and serum keratin 19 fragment were used to predict lymph node metastasis with an area under the ROC curve pair of 0.54, 0.46, 0.47, and 0.42, respectively. multivariate analysis revealed no independent predictors of lymph node metastasis among these preoperative parameters. Similarly, no preoperative predictors of tumor pathological invasion were identified for ROC curves or multivariate analyses.