Pulmonary dysfunction is a well-recognized complication of burn wound excision. It remains unclear whether this is caused by bacteria or inflammatory mediators released during excision of the wound. We produced a 15% full-thickness burn in 18 sheep, and between days 5 and 7 completely excised the wound under general anesthesia. Pulmonary parameters of static and dynamic lung compliance (CSTAT and CDYN), PaO2/FiO2, and pulmonary artery pressure (Ppa) were measured, as well as burn lymph, venous and aortic thromboxane B2 (TxB2), chemotactic index (chemotaxis/chemikinesis), and oxygen radical activity reflected in the level of lipid peroxidation in lung tissue. We noted a transient increase in burn lymph and venous TxB2 during excision, increasing from a preburn value of 200 and 220 +/- 50 pg/ml to 950 +/- 210 and 980 +/- 280 pg/ml, respectively. In 13 of 18 sheep, chemotactic activity and lung tissue lipid peroxidation, measured as malondialdehyde (MDA) content, were not increased. In this group only a very transient decrease in CDYN, PaO2/FiO2, and a 3 mm Hg increase in mean Ppa was seen with excision, with these parameters returning rapidly to baseline. Five of the 13 sheep had wound biopsy specimens that were greater than 10(6) organisms/gm tissue. In the remaining five sheep, plasma chemotactic index was also significantly increased with excision, as was lung MDA content, while decreases in CDYN, CSTAT, and PaO2/FiO2 and an increase in Ppa were more protracted. Three of these five sheep had wound biopsy specimens greater than 10(6) organisms/gm. We conclude that a release of thromboxane occurs during excision, which corresponds in time to transient lung dysfunction. If there is also a release of chemotactic factors, a more protracted pulmonary response occurs with evidence of O2 radical-induced lung changes.