Background: Although gastric cancer (GC) with para-aortic lymph node (PAN) metastasis is diagnosed as stage IV and basically treated with chemotherapy, it can be the candidate of conversion surgery when PAN metastasis was limited in localized area as oligometastatic status. Recently, we reported that pathological complete response at PAN (pCR at PAN) after chemotherapy was a good surrogate marker in conversion surgery for GC with PAN metastasis. However, it is difficult to predict pCR at PAN before surgery. Computed tomography (CT) is a standard procedure for clinical diagnosis of lymph node metastasis in GC.This study aimed to evaluate the reliability of CT to predict pCR at PAN after chemotherapy. Methods: The study examined patients who received radical gastrectomy with D2 and PAN dissection after chemotherapy for gastric cancer with PAN metastasis (#16a2/b1) from 2004 to 2020. PAN greater than 10mm in long axis was diagnosed as clinically positive. Patients were classified into 2 groups as pPAN-CR and pPAN-nonCR according to the pathological diagnosis of PAN. Accuracy, sensitivity, specificity, positive predictive value, and negative predictive values for PAN diagnosis were retrospectively investigated. Moreover, the relationship between shrinkage ratio at PAN and pathological diagnosis was also investigated. Results: A total of 46 patients were examined in the present study. The number of pPAN-CR and pPAN-nonCR patients were 31 and 15, respectively. 5-year overall survival rate was 76% in pPAN-CR and 0% in pPAN-nonCR patients. The accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of PAN were 52.2%, 60.0%, 48.4%, 36.0%, and 28.6%, respectively. Median shrinkage ratio of PAN was 38% (-20%-88%) in pPAN-CR and 40% (-20%-75%) in pPAN-nonCR patients. There was no statistically significant relationship between shrinkage ratio and pathological diagnosis of PAN (p=0.655). Conclusions: CT diagnosis for PAN after chemotherapy was not reliable. The indication of conversion surgery for GC with PAN metastasis should not be determined based on the size and shrinkage of PAN.