During the course of one year, 47 critical care patients with gram-negative bacillary pneumonia at Millard Fillmore Hospital were randomly assigned to aztreonam or tobramycin therapy (two to one). Of these, 40 were fully evaluable for microbiologic and clinical response. All evaluable patients had gram-negative organisms in tracheal aspirate culture specimens and confirmed susceptibility of the organism to both study drugs. There was no difference between the two groups with respect to the percentage of patients who received concurrent antibiotics for gram-positive organisms. More than 60 percent of the patients received mechanical ventilation. Essentially, all had new lung infiltrates as shown by chest radiography, leukocytosis, recent onset of fever, and increased volume of purulent secretions. Half had multilobar pulmonary infiltrates. Their mean age was 73 years, with none under age 50. Most had chronic obstructive pulmonary disease, congestive heart failure, or both. By the prognostic nutritional index criteria, over 70 percent were nutritionally deficient at entry. The majority of infections were caused by Pseudomonas, Enterobacter, Klebsiella, and Escherichia coli. Aztreonam eradicated 92 percent of the causative gram-negative organisms, compared with 57 percent for tobramycin (p less than 0.05). Aztreonam produced a favorable clinical response (cure or improvement) in 93 percent of patients, compared with 50 percent for tobramycin (p less than 0.05). There were no differences in the minor adverse effects observed in the two treatment groups. Overall, aztreonam was superior to tobramycin for treatment of pneumonia due to susceptible gram-negative bacteria in these critical care patients.