The anti-inflammatory/immunoparalytic phase of the systemic inflammatory response syndrome (SIRS) following major insult (surgery, thermal/traumatic injury) is of major clinical importance in the neonate, during which the risk of infection is particularly great. Here, the mechanisms by which TNF-a production is suppressed in response to infection are largely unknown. We questioned whether TNF-a itself could be a critical mediator of this suppression. Monocytes, isolated from cord blood (n=3), were treated with LPS (100 ng/ml), TNF-a (10 ng/ml, +/À anti-TNF-a antibody) for 18 and 36 h. Cells were then restimulated with LPS (Gram Àve) or Pam-3-Cys (Gram +ve) for 24 h. This was also done in the presence of selective inhibitors of MAP kinases p38, MEK and JNK. TNF-a, IL-6, IL-10 and IL-8 were quantified by ELISA CD86 and HLA-DR expression were determined flow cytometrically. Cells stimulated with LPS for 24 h produced TNF-a (282 pg/ ml), IL-10 (1,236 pg/ml), IL-6 (2,694 pg/ml) and IL-8 (2,144 pg/ml). In cells pre-exposed to TNF-a for 36 h, there was a significant suppression in TNF-a and IL-6 levels (9 and 221 pg/ml, respectively) (P<0.05) with minimal impact on IL-10 (1,206 pg/ml) and IL-8 levels (1,886 pg/ml). A similar effect was seen with Pam-3-Cys with a tenfold decrease in levels of TNF-a and IL-6 (86 fi 8.5 pg/ml and 458 fi 46 pg/ml, respectively) with no effect on IL-10 and IL-8 levels. Anti-TNF-a antibody negated this effect. Inhibition of p38 kinase reversed the TNF-a effect. Inhibition of the JNK and MEK kinases had no effect. A reduction in the expression of CD86 and HLA-DR was observed. This ex-vivo model of non-septic SIRS demonstrates that TNF-a, released during a major insult, can suppress subsequent monocyte responses to bacterial agents through p38 MAP kinase, making it a potential therapeutic target.