One of the problems of prolonged ventilatory therapy in acute respiratory failure (ARF) is the need to choose between tracheostomy after 48 to 72 hours of translaryngeal (TL) tracheal intubation or the continuous use of the TL tube for a period of 10 days. Too often the choice has been based on retrospective studies or personal preference. To investigate this problem prospectively, 52 adults in ARF were divided sequentially into 2 groups on their 3rd day of TL intubation. Patients in group I (G-I) retained the TL tube for a total of 11 days; those in group II (G-II) were tracheostomized on the 3rd day. The following factors ere used to evaluate the efficiency and complications in each group: patient's epidemiologic variables, daily pulmonary functions, severity of respiratory infections, and scores of post-intubation airway lesions. No consistent statistically significant differences between the two procedures were seen in the pulmonary functions or the range of individual patient variables. However, with an early tracheostomy, there was an eightfold greater incidence of contamination of the airway by new organisms, airway lesions were more frequent and severe, and the need for the tracheal tube was extended. To identify the epidemiologic variables and the pulmonary functions that discriminate between patients with serious airway lesions and those with mild lesions, and to evaluate the ability of these variables to differentiate the patients who died from those who survived, the distribution of all factors was compared in the two categories. The epidemiologic variables separated the patients according to their airway lesions only, while the difference in pulmonary functions was statistically significant only between the patients who died and those who survived.