Introduction: Esophageal cancer is an aggressive disease and is the eighth cause of malignant tumors worldwide. To treat dysphagia, auto expandable prosthesis (AEP) are used in order to optimize the nutritional status and quality of life of the patients.Objectives: To quantify patients’ mortality and to evaluate variables related with this outcome.Materials and methods: Retrospective cohort study that involved 135 patients with esophageal malignant obstruction that required AEP insertion. Survival and incidence density rates were estimated. The effect of these variables on the probability of death was assessed by using Cox models.Results: Overall mortality rate was 13.7 deaths per 100 patients/month (95%CI: 10.9-17.1). Univariate analysis showed significant differences in survival functions according to pre-intervention albumin levels and the prosthesis size (>12cm). In the Cox model, albumin level (HR=0.53, 95%CI 0.31 to 0.89) was the only significant result.Conclusions: AEP represent a therapeutic option to improve symptoms in patients with advanced esophageal and gastroesophageal junction tumors. This technique has few complications and its clinical success is around 90%. Patients’ nutritional status and length of the stenosis caused by the tumor are variables that must be evaluated before performing a procedure as they seem to be related to mortality.