Discussion of paranasal sinusitis as a nosocomial infection in the mechanically ventilated intensive care (ICU) patient has recently been intensified. Some authors have emphasized nasotracheal intubation as a possible pathogenetic pathway. The aim of this study was to investigate the impact of nasotracheal or orotracheal intubation on the development of sinusitis in ICU patients. METHODS In a prospective study, we followed 44 patients who required mechanical ventilation (greater than 24 h) in the ICU because of prolonged recovery from abdominal, thoracic, or posttraumatic surgery. Twenty patients were intubated nasotracheally and 24 orotracheally. Assignment to the groups was random. All were provided with a nasogastric tube and initially treated with systemic antibiotics. They received local antimicrobial prophylaxis of the nose, oropharynx, and stomach. Daily a-scan examinations of the maxillary sinuses were performed from the day of admission to the ICU until extubation, tracheotomy, death, or transfer. The average observation period was 6.9 days in the oral group and 7.1 days in the nasal group. In the case of a pathologic finding, aspiration of the antral sinus was carried out. In this study sinusitis indicated a sonographic finding; it did not necessarily imply a bacterial infection. RESULTS At the beginning of the observation period, 6 patients in the oral group and 4 in the nasal group already had a pathologic maxillary sinus finding. At the end, in 15 of 24 in the oral group and 19 of 20 in the nasal group unilateral or bilateral sinusitis could be demonstrated. Development of bilateral sinusitis (13/20 in the nasotracheal group, 8/24 in the orotracheal group) was mainly observed after the appearance of unilateral sinusitis. The site corresponded to the site of the nasal tube in 65%. Unilateral paranasal infection was observed in nasotracheally and orotracheally intubated patients after an average of 2.8 and 2.6 days, respectively, whereas bilateral sinusitis had an average time delay of 4.5 and 5.7 days. Aspiration of the maxillary sinus was performed in 22 of 34 cases with sinusitis. Pathogenic organisms could be demonstrated in 7 of 13 nasotracheally intubated patients but only 2 of 9 with orotracheal tubes. CONCLUSIONS We found that patients intubated orotracheally developed significantly less sinusitis than those intubated nasotracheally. Edema, local infection of the nasal mucosa, or mechanical obstruction of sinus drainage pathways by the tube are possible explanations. The fact that 63% of orally intubated patients had a pathologic maxillary sinus finding as well suggests that in addition to other reasons, an increased central venous pressure, positive pressure ventilation, and the supine position must be regarded as predisposing factors that increase the incidence of sinusitis. We conclude that the conditions of critically ill patients predispose to the development of sinusitis. Nasotracheal intubation is to be regarded as an additional risk, and therefore oral intubation should be preferred.