SUMMARY. Histopathological tumor regression grade (TRG) has been shown to be a prognostic factor in patients with esophageal cancer after neoadjuvant radiochemotherapy (RCT). The system introduced by Mandard to group TRG (Cancer 1994;73:2680–2686) has been used to analyse and discuss its prognostic significance on survival in a single institution retrospective analysis: TRG 1 (complete regression) – TRG 5 (absence of regressive changes). Sixty patients with locally advanced (T3/4 or N1) adenocarcinoma or squamous cell carcinoma received cisplatin-based RCT. Three to four weeks later operation for curative intent was performed. Median follow-up was 17.7 months. Histopathological tumor stages were stage 0 in 17%, stage I in 10%, stage II in 60%, stage III in 12% and stage IVA in 1%. The 5-year overall survival (OS) rate was 35%. In univariate analysis, ypN-status and TRG correlated significantly with OS (P = 0.004, P = 0.0008, respectively). While OS of TRG 1 differed significantly from all other groups, no differences in OS between the other TRG groups were seen. Patients with complete tumor regression after neoadjuvant RCT showed a much better survival than patients with tumors that responded less to induction therapy. Further qualitative subdivision of tumor regression could not identify patient groups with significant differences in prognosis. After comparing our data with the literature, it is reasonable to consider classifying all patients into ‘Complete tumor regression’ and ‘Incomplete tumor regression’.