The incidence of esophageal cancer still increases in many countries. Overall prognosis is poor despite advances in surgical and multimodal therapy. Lymphatic spread per se is a strong prognostic factor but some newer data proposed a better prognostic value of the lymph node ratio (LNR). The role of neoadjuvant chemoradiation (neoCRTx) is still discussed controversially. We evaluated the long-term outcome after resection of esophageal cancer in 221 patients (58% after neoCRTx).Methods: We analyzed the outcome in 221 patients (83% male, median age 60 years) who underwent esophagectomy since 1988. Initial staging consisted in CT-scans and endosonography, 59% had squamous cell carcinoma and 41% adenocarcinoma. Tumors were in the lower third in 60% and in the middle/upper third in 33% / 7%. NeoCRTx (36 Gy+FU+Cisplatin) was applied in 58% (indications >T2 or node positive). At initial staging 52% of all patients were staged node-positive (cN+). In the group undergoing neoCRTx 68% were initially cN+ whereas in the patients without neoCRTx only 29% were cN+. Survival was analyzed by the Kaplan-Meier- and Cox-models.Results: After neoCRTx 99 of 128 patients (77%) showed partial or complete remission. Free resection margins were achieved in 91 %. At pathological assessment 47 % of all patients were node-positve (pN+). Compared to initial staging, the final nodal status increased from 29% (cN+) to 47% (pN+) in patients without neoCRTx but decreased from 68% (cN+) to 43% (pN+) after neoCRTx. Actuarial five-year survival was 28%. Univariately five-year survival was better in patients with high differentiation (32% in G1/2 vs 18% in G3/4; p<0.01), with negative LN (47% vs 10% in pN+; p<0.001), with a LNR ≤ 0.1 (42% vs 5% in LNR > 0.1; p<0.001), with response to neoCRTx (43% vs 16% in pts. without CRTx or response; p<0.001) or with free margins (34% vs 12% in R+; p=0.001). The number of examined LN and histological type did not influence survival. Multivariate analysis revealed good differentiation (p<0.03), neoCRTx (p<0.03) and a LNR < 0.1 (p<0.001) as independent prognostic factors. The resection margin was not a significant factor in the Cox-model. This might be explained by the low rate of positive margins after neoCRTx (2% vs 14% without CRTx; p<0.01) and the low total number.Conclusions: Initial evaluation understaged nodal disease. The LN-ratio is the strongest predictive factor after resection of esophageal cancer. The effect of neoadjuvant chemoradiation on survival was mainly due to an increase in the rate of negative resection margins and to the eradication of tumor cells in initially positive LNs.