BACKGROUND: Hypothermia is known to suppress inflammation in various experimental and clinical settings. We wanted to investigate how the suppressed inflammation by hypothermia is affected during rewarming. METHODS: Mice were being assigned to normothermia (37degrees C) or hypothermia (32degrees C). After 30 minutes at the assigned temperature, lipopolysaccharide was administered intratracheally. The mice were then randomly grouped and subjected to 4 hours of normothermia (N), 24 hours of normothermia (NN), 4 hours of hypothermia (H), or 4 hours of hypothermia followed by normothermia for the next 20 hours (HN). In another experiment, other HN mice were treated with varying doses of anti-TNF-alpha or anti-IL-1beta antibodies (0, 6.25, 12.5, 25, and 50 microg/250 microl) immediately prior to rewarming. RESULTS: The neutrophil counts of BAL fluid (x104/ml) were 23.0 +/- 13.1 in the N, 6.4 +/- 3.1 in the H (p = 0.002 vs N), 20.4 +/- 10.2 in the NN, and 49.7 +/- 21.0 in the HN (p = 0.005 vs H; p < 0.001 vs NN). Myeloperoxidase activity of the lung (unit/microg) was 6.7 +/- 2.9, 7.9 +/- 1.9, 17.8 +/- 4.0 (p < 0.001 vs N), and 12.9 +/- 5.9 (p = 0.034 vs H, p = 0.028 vs NN), respectively. Compared with control HN, total WBC and neutrophil counts of mice treated with anti-TNF-alpha antibody or anti-IL-1beta antibody prior to rewarming were lower at all tested doses. The combination of both anti-TNF-alpha or anti-IL-1beta antibodies was not increasingly reducing the neutrophilic sequestration. CONCLUSIONS: Rewarming from induced hypothermia resulted in augmentation of neutrophilic sequestration of endotoxin-injured lung. Treatment with antibodies against TNF-alpha or IL-1beta prevented this rebound of neutrophilic infiltration.