OBJECTIVE Neoadjuvant chemotherapy (nCT) followed by surgery is one of the standard treatments for resectable esophageal squamous cell carcinoma (ESCC). According to the Response Evaluation Criteria in Solid Tumors, endoscopic evaluation of a primary lesion is not recommended during nCT because of reduced objectivity. This study aimed to develop and validate endoscopic evaluation criteria for nCT. METHODS This study retrospectively investigated patients with T2/3 ESCC who underwent nCT followed by radical esophagectomy across two institutions (test and validation sets). We retrospectively estimated the therapeutic effect by classifying patients according to degree of tumor shrinkage (evaluated with endoscopy) as follows: marked reduction (MR), half reduction (HR), insufficient reduction (IR), and progressive disease (PD). Three endoscopists evaluated patients in the test set. Another three endoscopists evaluated patients in the validation set. We analyzed recurrence-free survival (RFS) 3 years after surgery. RESULTS Of 129 patients in the test set, 44 had MR, 35 had HR, 44 had IR, and six had PD. The 3-year RFS rates were 55% (overall), 79% (MR), 54% (HR), 35% (IR), and 33% (PD). Of 91 patients in the validation set, 22 had MR, 49 had HR, 18 had IR, and two had PD. The 3-year RFS rates were 54% (overall), 77% (MR), 55% (HR), 22% (IR), and 50% (PD). CONCLUSIONS Our endoscopic criteria for nCT predicted prognosis; however, future studies are needed to further investigate our criteria before general application in the clinical setting.