This study aims to evaluate the accuracy of endoscopic ultrasound (EUS) in the staging of esophageal cancer after neoadjuvant chemotherapy (NAC). Articles were searched in Medline, Pubmed, Cochrane Database of Systemic Reviews, Google scholar, and EMBASE. Two reviewers independently searched and extracted data. Meta-analysis of the accuracy of EUS was analyzed by calculating pooled estimates of sensitivity, specificity, likelihood ratios (LR), and diagnostic odds ratio (DOR). Pooling was conducted using either fixed-effects model or random-effects model depending on the heterogeneity across studies. Sixteen studies (n = 724) were included in this analysis. The pooled sensitivity and specificity of EUS to diagnose T1 stage tumor was 23% (95% confidence interval [CI] 16-32%) and 95% (95%CI 93-97%), respectively. For T2 stage, EUS had a pooled sensitivity and specificity of 29% (95%CI 19-41%) and 84% (95%CI 77-88%). The pooled sensitivity and specificity of EUS were 81% (95%CI 72-88%) and 42% (95%CI 33-52%) in determining T3 stage tumor. To diagnose T4 stage tumor, EUS had a pooled sensitivity of 43% (95%CI 31-56%) and specificity of 96% (95%CI 94-97%), respectively. In determining N stage, the pooled sensitivity and specificity of EUS were 69% (95%CI 58-79%) and 52% (95%CI 42-62%). EUS is a moderately accurate technique in staging esophageal cancer after NAC. Its sensitivity is relatively high in T3 while specificity is high in other T stages (T1, T2, and T4). Tumors restaged by EUS as T4 should not be assigned to surgery because it is very likely to be inoperable. EUS is not reliable for N staging with its poor sensitivity and specificity. Subgroup analysis shows that staging accuracy did not improve with time.