While open window thoracostomy (OWT) is a safe procedure and is indicated in patients who have thoracic empyema either with or without a bronchopleural fistula, it may prolong the hospital stay. We retrospectively analyzed the relationship between the etiology of thoracic empyema and the open window interval (OWI). Between January 1986 and May 1997, 53 patients resistant to conventional therapy underwent OWT for thoracic empyema at the Department of Surgery of the National Minami-Fukuoka Chest Hospital. The patients were classified into five groups based on the etiological findings of thoracic empyema. 44 patients also underwent closure of the window until June 1999. The average OWI was 180.4 +/- 51.9 (mean +/- SE) days for postoperative empyemas in lung cancer, 128.0 +/- 32.1 days for bacterial nontuberculous empyemas, 189.6 +/- 24.1 days for fungal empyemas, 365.8 +/- 201 days for empyemas caused by atypical mycobacteria and 322.0 +/- 58.7 days for tuberculous empyemas. There was no evidence that the OWI was related to either sex, age, etiology of thoracic empyemas, performance status, the existence of bronchopleural fistulae, complications of diabetes mellitus or preoperative malnutrition status in multivariable models. 5 patients underwent a second OWT because of recurrence of empyema. Mortality rate was 7.5%. There was no relationship between clinical factors including nutritional assessment and OWI. OWT generally is a safe and effective procedure for thoracic empyema resisting to conventional therapy except that it can make an extended hospital stay necessary. Copyright 2001 S. Karger AG, Basel