Background Pathologic response to neoadjuvant chemotherapy (neoCTX) is a prognostic factor in many cancer types, and early prediction would help to modify treatment. In patients with gastric and esophagogastric junction (AEG) cancer, the accuracy of FDG PET-CT to predict early pathologic response after neoadjuvant chemotherapy (neoCTX) is currently not known. Methods From a consecutive cohort of 72 patients, 44 patients with resectable, locally-advanced gastric cancer or AEG Siewert type II and III received neoCTX after primary staging with endoscopic ultrasound, PET-CT and laparoscopy. Overall, 14 patients did not show FDG uptake, and the remaining 30 were restaged by PET-CT 14 days after the first cycle of neoCTX. Metabolic response was defined as decrease of tumor standardized uptake value (SUV) by greater than or equai to35%. Major pathologic regression was defined as less than 10% residual tumor cells. Results Metabolic response after neoCTX was detected in 20/30 (66.7%), and non-response in 10/30 (33.3%) patients. Among metabolic responders, n = 10 (50%) showed major and n = 10 (50%) minor pathologic regression. In non-responders, n = 9 (90%) had minor and 1 (10%) a major pathologic regression. This resulted in a sensitivity of 90.9%, specificity 47.3%, positive predictive value 50%, negative predictive value 90% and accuracy of 63.3%. Conclusion Response PET-CT after the first cycle of neoCTX does not accurately predict overall pathologic response. However, PET-CT reliably detects non-responders, and identifies patients who should either immediately proceed to resection or receive a modified multimodality therapy. Trial registration The trial was registered and approved by local ethics committee PB_2016-00769. Keywords: Histopathologic regression, PET-CT, Gastric cancer, AEG