With the inhalation of smoke, there are both cardiopulmonary changes and elevated levels of carbon monoxide (CO). We hypothesize that these changes in cardiopulmonary function are the result of a histotoxic hypoxia associated with CO poisoning. This hypothesis was tested in chronically instrumented sheep (n = 19). Piezoelectric crystals were attached to the left ventricle for the measurement of its external minor and major diameters in addition to wall thickness. A pressure transducer was placed in the left ventricle via the apex. The caudal-mediastinal lymph node was also cannulated. After a five-day recovery period, six sheep (smoke group) were insufflated with four series of 16 breaths (700 ml/breath) of cotton smoke, and five sheep (control group) were insufflated with air using a modified bee smoker (smoke group: COHb, 90 +/- 6%; control group: COHb, 6 +/- 1%). Eight sheep (CO group) were ventilated with 2% CO in air to reach a COHb of 90% (COHb, 92 +/- 1%). In the smoke group, lung lymph flow reached 42 +/- 10 ml/hr at 24 hours after smoke insufflation (baseline, 6 +/- 1 ml/hr). The maximum elastance of the left ventricle (end-systolic pressure-volume ratio), a sensitive index of myocardial contractility, was significantly decreased from a baseline of 6.5 +/- 0.9 to 3.3 +/- 0.7 mm Hg/ml. In the control and CO group, neither lung lymph flow nor maximum elastance varied from the baseline value. We conclude that the cardiopulmonary dysfunction after smoke inhalation does not occur after a similar exposure to CO. Initial CO poisoning alone is not a causative factor of cardiopulmonary dysfunction after smoke inhalation.