Studies with ex vivo stimulation of whole blood samples from injured patients have revealed a diminished production capacity for a broad range of secretory products, including inflammatory cytokines. Recent interest has focused on the release of mediators in serum that depress the cell-mediated immune response following trauma. The involvement of the lipid mediator prostaglandin E2 (PGE2) has been assumed because it is a potent endogenous immunosuppressor. In the present study, we tested the hypothesis that inhibitory substances circulating in the patient's serum after a major musculoskeletal trauma might impair leukocyte function by evaluating the effect of such serum on cytokine release in a whole blood model. Six females and three males undergoing elective total hip replacement were included in the study. Ex vivo LPS-induced TNF-alpha and IL-10 were measured in whole blood sampled preoperatively and added serum taken before, at the end of operation, and at postoperative days 1 and 6 with saline as negative control. LPS induced significant releases of TNF-alpha and IL-10 in whole blood. Addition of preoperative, postoperative, and day-1 postoperative serum did not alter the LPS-induced release of TNF-alpha as compared to saline. In the presence of serum from postoperative day 6, however, the expression of TNF-alpha was significantly reduced as compared to saline and preoperative serum (p = 0.021 and 0.008, respectively). Neither of the serum samples altered the release of IL-10. PGE2 was significantly (p = 0.008) increased in serum at postoperative day 6 as compared to preoperative levels. In conclusion, these data show that at day 6 after major orthopaedic surgery, the patient serum contained activity that inhibited ex vivo LPS-induced TNF-alpha release. The potent TNF-alpha inhibitory activity found at day 6 after injury correlated with increased levels of PGE2 and indicates cell-mediated hyporesponsiveness to a second stimulus.