BACKGROUND: Multimodality treatment that includes esophagogastrectomy may represent the best option for curing accurately staged patients with esophageal cancer. We analyzed the impact of incomplete resection on outcomes after esophagogastrectomy for esophageal cancer. METHODS: The incidence of positive margins for patients who underwent esophagogastrectomy without induction therapy for pathologic T1-3N0-1M0 esophageal cancer of the mid and lower esophagus from 2003 to 2006 in the National Cancer Database was analyzed with multivariate logistic regression. The impact of positive margins on survival was assessed using Kaplan-Meier and Cox proportional hazards analysis. RESULTS: Positive margins occurred in 342 of 3,125 patients (10.9%) who met study criteria. Increasing clinical T status was an independent predictor of positive margins in multivariate analysis, but the chance of positive margins decreased with larger facility case volumes. The presence of clinical nodal disease was not predictive of an incomplete resection. The 5-year survival of patients with positive margins (13.8%, 95% confidence interval [CI]: 10.5% to 18.1%) was significantly worse than that for patients with negative margins (46.3%, 95% CI: 44.4% to 48.3%, p < 0.001). Both microscopic residual disease (hazard ratio 1.37, 95% CI: 1.16 to 1.60, p < 0.001) and gross residual disease (hazard ratio 1.98, 95% CI: 1.62 to 2.42, p < 0.001) predicted worse survival in multivariate analysis of the entire cohort. Receiving adjuvant chemoradiation therapy slightly improved 5-year survival of patients with positive margins (16.9%, 95% CI: 11.3% to 23.6%, versus 13.5%, 95% CI: 9% to 20.3%, p < 0.001). CONCLUSIONS: Positive margins are associated with poor survival, and adjuvant therapy only marginally improved prognosis. Future studies are needed to better evaluate whether induction therapy can lower the incidence of positive margins.