The postoperative mortality after esophagectomy still remains a major factor influencing the prognosis of esophageal cancer and largely depends on the patient's preoperative physiological status. A composite scoring system was developed to predict the risk of esophagectomy, based on quantitative assessment of preoperatively available physiological parameters. The scoring system was reviewed retrospectively on operated patients and evaluated prospectively in two subsequent patient groups. An initial retrospective multivariate analysis of 432 esophagectomy patients identified a compromised general status (p = 0.001) and poor cardiac (p < 0.001), hepatic (p < 0.05), and respiratory (p < 0.05) functions as independent predictors of a fatal postoperative course. Based on the relative risks associated with individual impaired organ functions--general status 3.6, cardiac function 2.8, hepatic function 2.1, pulmonary function 1.7--a composite risk score was established. A prospective study in 121 patients confirmed that this composite scoring system provides better identification of high-risk patients than does any of the individual parameters alone. Including this composite score into the process of patient selection and choice of procedure resulted in a decrease of postoperative mortality from 9.4% (52/553) to 1.2% (4/323) (p = 0.001). The risk of death after esophagectomy for esophageal cancer can be objectively assessed prior to surgery and quantified by a composite risk score. This score provides a useful tool in refining the criteria of patient selection for resection and choice of procedure, and markedly reduces postoperative mortality when applied prospectively.