Arteriovenous differences in lactate (AVLAC) across the lungs are usually small and close to zero. However, it has recently been reported that the lungs can produce increased amounts of lactate in some patients with acute respiratory distress syndrome (ARDS). The aim of this study was to evaluate lactate production in various types of acute lung injury requiring mechanical ventilation and hemodynamic monitoring. Since the differences involved are usually small, minor errors in lactate measurement could greatly influence AVLAC. Based on an analysis of these errors (see text for details), we averaged five arterial and venous samples for each measurement. We investigated 122 patients: 43 with acute lung injury (ALI), nine with cardiogenic pulmonary edema (CPE), 37 with bronchopneumonia (BPN), seven with single lung transplantation (LTX), and 26 with other causes of respiratory failure (OTHER). There was no difference in arterial lactate between the various groups. AVLAC was higher in patients with ALI than in the other groups (0.20+/-0.23 versus 0.07+/-0.11 mEq/L). In patients with ALI, AVLAC was proportional to the Murray's lung injury score (-0.032+/-0.032x; r = 0.46, p < 0.01). Lung lactate production was calculated as the product of the cardiac index times AVLAC and was significantly higher in patients with ALI than in the other groups (0.69+/-0.88 versus 0.19+/-0.30 mEq/min; p < 0.05). In patients with ALI, lung lactate production was inversely related to the PaO2/FIO2 (1.42 - 0.005x; r = 0.35, p < 0.05) but directly related to the venous admixture (-0.36 + 0.003x; r = 0.49, p < 0.01) and the lung injury score (-0.19 + 0.36x; r = 0.45, p < 0.01). Lung lactate production was not significantly related to arterial lactate levels. These data indicate that AVLAC and lung lactate production can be increased in patients with ARDS but remain within the normal range in other types of respiratory failure.