A prospective, double-blind, placebo-controlled study was conducted to determine the safety and efficacy of endotracheal tobramycin (ETT) for treatment of gram-negative bacterial pneumonia. Patients were randomized to either 40 mg of tobramycin or a placebo instilled endotracheally every 8 h. Patients also received intravenous tobramycin plus either cefazolin or piperacillin. Of 85 patients enrolled, 41 were assessable. Most microbiologic diagnoses were made by endotracheal aspiration with strict grading criteria. The clinicalradiographic responses of patients and standard demographic data were recorded. Pseudomonas aeruginosa, "multiple pathogens," and KkbsieUa-Enterobacter-Serratia-Citrobacter species were isolated in 41, 32, and 15% of the instances, respectively. Causative pathogens were eradicated from sputum significantly more frequently by patients who received ETT (P < 0.05). However, no significant differences were noted in the clinical outcomes of the two study groups. No local adverse reactions attributable to the administration of this agent were observed, but four patients had supraventricular tachycardia, compared with none who received the placebo (P = 0.053). ETT may be considered as adjunctive therapy for seriously ill individuals.Gram-negative bacterial pneumonia is known to be associated with substantial morbidity and mortality (1,4,6,18). Despite broad clinical experience with this disease, there is no consensus about issues as basic as the optimal number of antibiotics for therapy or the length of treatment. Although some researchers have recommended two-drug parenteral therapy for most cases (19), recent data have demonstrated the efficacy of newer P-lactam antibiotics as monotherapy for some patients (1,13). Similarly, the role of endotracheal or aerosolized antibiotics as either sole or adjunctive therapy for gram-negative bacterial pneumonia treatment or prevention has been controversial. Most interest has focused on aminoglycosides because of perceived problems with their efficacy when they are administered only systemically (5,19). The appeal of endotracheally administered antibiotics lies in the ability to place the antibiotic directly at the site of infection while potentially avoiding risks of systemic toxicity. Klastersky and colleagues have demonstrated the apparent benefit of prophylactic endotracheal gentamicin in the prevention of gram-negative bacterial lower-respiratory infections in patients with tracheostomies (10). His group has also shown the benefit of endotracheal gentamicin over the systemically administered drug in the management of small numbers of patients with gram-negative bacterial lower- respiratory infections and has correlated outcomes with antibiotic levels within endobronchial secretions (9). However, endobronchial administration of these agents has not been widely accepted, in part because of concerns regarding adverse reactions and colonization with resistant organisms. Double-blind, placebo controlled, prospective, randomized studies to assess the role of...