Pulmonary edema following smoke inhalation is due to the chemical toxins in smoke and not to the heat. We have shown that acrolein, a common component of smoke, induces pulmonary edema, perhaps via release of leukotrienes. We, therefore, hypothesized that acrolein, a component of smoke from burning cotton, might have a major role in producing pulmonary edema in sheep after cotton smoke inhalation and that BW-755C, a combined cyclo- and lipoxygenase inhibitor, would prevent the edema, whereas indomethacin, a cyclooxygenase inhibitor, would not. In control anesthetized sheep (n = 7), 128 breaths of cotton smoke induced no change in pulmonary arterial pressure but induced increases (P < 0.05) in pulmonary lymph flow from 4.4 +/- 0.8 (SE) to 15 +/- 2.7 ml/h, lymph protein flux from 0.25 +/- 0.08 to 0.80 +/- 0.16 g/h, and blood-corrected wet-to-dry weight ratios from a normal value of 3.8 +/- 0.07 (n = 9) to 4.5 +/- 0.18. Indomethacin (n = 6) did not significantly prevent these changes, whereas BW-755C decreased lung lymph flow change from 5 +/- 1 to 7 +/- 2 ml/h (P = NS), lymph protein flux from 0.25 +/- 0.08 to 0.35 +/- 0.1 g/h (P = NS), and weight-to-dry ratio from normal to 3.9 +/- 2.1 (P = NS). These data suggest leukotrienes may have a role in producing cotton smoke-induced noncardiogenic pulmonary edema.