Considerable controversy exists over the use of furosemide for the prevention or treatment of posttraumatic respiratory insufficiency. The conflict revolves around the use of a diuretic in a patient with this condition. There is some evidence that a nondiuretic effect of furosemide may be responsible for the reported improvement in lung function. We studied the response to furosemide of the pulmonary microvascular fluid filtration rate reflected in lung lymph flow (Qlym) in the normal lung. Using the unanesthetized sheep lung lymph preparation of Staub, we found a 30% decrease in Qlymafter 80 mg furosemide. However, the majority of the decrease occurred within 15 min after injection when diuresis was just beginning. This response appeared to be due to a large decrease in pulmonary venous resistance, decreasing hydrostatic pressure. Protein flow (QlymX lymph protein content) remained constant. Pulmonary artery pressure remained constant with left atrial pressure decreasing slightly. We have demonstrated that in the normal lung, furosemide significantly decreases the fluid filtration rate by a nondiuretic effect. Further studies of this response should help resolve the controversy over the indications for diuresis and center more attention on the actual mechanism of action of furosemide.