BACKGROUND: Results are conflicting and no population-based studies are available regarding the postoperative mortality after intrathoracic anastomotic leakage. The current study addressed the unselected and independent fatality rate of intrathoracic esophageal anastomotic leaks after resection for cancer. METHODS: A prospective, nationwide study was conducted in Sweden in April 2001 through December 2005. Details concerning patient and tumor characteristics, surgical procedures, postoperative anastomotic leakage, and mortality were collected prospectively. Logistic regression was performed to estimate odds ratios (ORs) and 95% confidence intervals (95% CIs), adjusted for age, tumor stage, comorbidity, and hospital volume. RESULTS: Among 559 resected patients with an intrathoracic anastomosis, 44 patients (7.9%) sustained an anastomotic leak within 30 days of surgery. Of these, 8 patients (18.2%) died within 90 days of surgery, compared with 32 of the 515 patients without leakage (6.2%) (P = .003). The adjusted OR of postoperative death following intrathoracic anastomotic leakage was increased 3-fold compared with those without such a complication (OR 3.0, 95% CI 1.2-7.2). CONCLUSION: Intrathoracic anastomotic leakage after esophageal resection for cancer remains a major risk factor for short-term postoperative death in an unselected, population-based setting.