Introduction: Heat stress prior to endotoxemia attenuates cardiopulmonary injury and improves survival. The mechanism of protection is unknown. Our previous study showed that plasma TNF-α. IL-10, and nitric oxide production were similar in heated and normothermic endotoxemic rats while MIP-2 was elevated in the heated group. We investigated if pretreatment of macrophages with heat prior to exposure to LPS results in modulation of cytokine production. Methods: Mouse macrophage cells (RAW 264.7), were seeded in three 24 well plates and incubated at 37°C for 24 hrs. Plate 1 was kept at 37°C and 12/24 wells received E. Coli LPS (100 ng/ml) on day 3. Plate 2 was heated on day 3 (42°C for 30 min) and 6 hrs later, LPS was added to 12/24 wells. Heat shock protein was produced by this method in the past. Plate 3 was heated on days 1.2 and 3 and 12/24 wells received LPS. 6 hrs after the last heating. All plates were then incubated for 24 hrs at 37°C and their supernatants were removed for analysis of TNF-α,IL-1β, and MIP-2. Cell count and viability was determined. One way ANOVA was used to compare groups and the Student-Newman-Keuls post-hoc test was used when necessary. Results: Cells which were triple heated and then given LPS produced more MIP-2 than single heated cells or normothermic cells (figure). However, triple heated cells produced TNF-α compared to single heated cells receiving LPS (49,940+4800 vs. 64.520+8300 pg/ml per 100,000 cells, p<0.05) but a similar amount compared to unseated cells treated with LPS. Heated or normothermic cells without LPS did not produce any TNF-α. IL-1β or MIP-2. Cell viability was similar in all groups. Conclusions:. Heat stress may be protective in endotoxemia by increasing MIP-2 production. MIP-2 production is probably directly stimulated by LPS and is not totally dependent on TNF-α and IL-1β production. Repeated episodes of heat stress may have a different effect on cytokine production than a single episode of heat stress. The clinical significance of repeated heat stress needs to be determined.