Introduction: The "Fistula Day" multicontinent, multinational, multicentered project has revealed a 14.7 % mortality rate in patients assisted for gastrointrestinal fistulas (GIF) in Latin American and European hospitals. Mortality associated with GIF might be explained for the clinical-surgical condition of the patient, the operational characteristics of the hospital, and the surgical practices locally adopted in the contention, treatment and resolution of GIF. Objective: To assess the influence of surgical practices adopted in the hospital upon GIF outcomes. Study design: Cohort-type study. Three cross-sectional examinations were done during the completion of the exercises of the "Fistula Day" project: on admission in the study serie, and at 30 and 60 days after admission. Study serie: One hundred seventy seven patients (Males: 58.2 %; Average age: 51.0 ± 16.7 years; Ages ≥ 60 years: 36.2 %) assisted in 76 hospitals of Latin America (13 countries) and Europe (4). Methods: Surgical practices adopted in the management of GIF were documented such as the use of computerized axial tomography (CAT) and oral ingestion of contrast for examination of the fistula path, the use of open abdomen and devices for temporary closure of the abdominal wall, the administration of somatostatin and analogs for promoting the closure of the fistula, reoperation for fistula closure, and admission in the ICU. Results: Usage rate of surgical practices was as follows: CAT + oral use of contrast: 39.5 %; Use of open abdomen: 31.1 %; Use of somatostatin and analogs: 22.6 %; Admission in the hospital ICU: 31.6 %; and Surgery for GIF closure: 33.9 %; respectively. Surgical practices were more frequently used in the treatment and containment of enteroathmosferic fistulas (EAF). Surgical practices adopted by participating hospitals did not imply a higher rate of GIF closure, but were associated instead with a higher mortality and prolongation of hospital stay. Conduction of surgical practices was independent from the guidelines followed by the medical teams in the management of GIF. Availability of surgical practices, and access of medical teams to them, were independent from the operational characteristics of the surveyed hospital. It is to be noticed the existence of a hospital unit dedicated to intestinal failure translated to a lower use of the techniques for open abdomen and temporary closure of the abdominal wall, which, in turn, translated to a higher likelihood of GIF spontaneous closure. Conclusions: Currently, the adoption of surgical practices for containment and resolution of GIF does not result in a higher GIF closure rate. It is likely the existence of a hospital unit specialized in the management of intestinal failure might bring about a higher rate of non-surgical closure of GIF.