Background: While it is commonly reported that over half of patients with esophageal cancer (EC)initially present with systemic metastases, data from the American Cancer Society indicate that at least 55%, and as many as 70%, present with locoregional disease alone. While esophagectomy provides the highest probability of long-term cure, the perception of poor overall survival and high perioperative mortality results in many candidates never being referred for surgical evaluation. We hypothesized that esophagectomy for EC is underutilized and assessed the prevalence of resection in national, state and local cancer data registries. Patients and Methods: Clinical stage, surgical and non-surgical treatments, age, and race of patients with EC were identified from the Surveillance, Epidemiology and End Results (SEER) registry (1988-2004), the Hospital Association of NY State registry (HANYS 2007) and a single referral center (2000-2007). SEER identified a total of 25,306 patients with EC(average age: 65.0 years, male-to-female ratio: 3.1); treatment modality was identifiable in 22,617. HANYS identified 1519 admissions for cancer of the esophagus and cardia (average age: 67 years, M:F ratio 3:1); treatment modality was identifiable in all patients, though stage was not available from NY state registry data. A single referral center identified 420 patients (52/year; average age: 67 years, M:F, 3:1); treatment modality and stage were available in all. For SEER data, logistic regression was used to examine determinants of esophageal resection; variables tested included age, race and gender. Results: Sixty-three percent of the SEER population (14,240) was classified as having locoregional disease. Surgical resection was performed in 25% (5,644/22,617) of the total and only 39.6% of potentially resectable patients (5,644/14,240) . Similarly, resection was performed in 28.6% (443/1519) of the total admissions from NY State. By comparison, 65% (272/420) of patients at a specialized referral center underwent surgical resection. Fifty seven percent (8,133) of patients in the SEER registry received no operative therapy. Resection rates for EC did not change between 1988 and 2004. Males were more likely to receive operative treatment (OR: 1.1, p<0.05). Whites and blacks were less likely to undergo surgery than Asians (OR: 1.47, p<0.01). Conclusion: Surgical resection for locoregional EC is likely underutilized. Racial variations in esophagectomy are significant. While the pattern of utilization is poorly documented and largely unexplained, we believe that referral to specialized centers may result in an increase in patients considered for surgical therapy.