Background: In primary esophageal cancer, studies have frequently focused on surgical patients in an effort to link outcome to hospital-or surgeon-related experience, with operative mortality used as the main outcome measure. Many studies have found an inverse relationship between operative mortality and hospital volume and surgical expertise. This study aims to assess the influence of surgeon-related expertise and hospital volume on the relative survival of operated esophageal cancer patients.Methods: From January 1994 to January 2002, a total of 1149 consecutive patients with primary esophageal cancer were diagnosed in the region of the Comprehensive Cancer Center North-Netherlands. As a proxy for surgeon-related expertise, hospitals in this region were categorized into three types: university, teaching nonuniversity, and nonteaching hospitals. The influence of hospital type on the relative survival of operated patients was studied by a multivariate Poisson regression model.Results: Of the 1149 patients, 18.5% underwent surgery. There was no evidence of selective referral for surgery between the three hospital types with regard to age, tumor stage, and location. For operated patients, the 5-year relative survival was 49.2% for the university hospital versus 32.6% and 27.3% for teaching nonuniversity and nonteaching hospitals, respectively (P = .0039). When adjusted for age, tumor stage, hospital volume and referral frequency, the relative excess risk of death for the university hospital was considerably lower at .57 (95% confidence interval, .29-1.12) compared with nonteaching hospitals and .43 (95% confidence interval, .24-.76) compared with teaching nonuniversity hospitals (P = .0126).Conclusions: In our region, patients with esophageal cancer who underwent esophagectomy in the university hospital had a markedly better relative survival compared with those who underwent surgery at teaching nonuniversity and nonteaching hospitals, emphasizing the need for referral of esophageal surgery to centers with a greater experience.