In a previous study we demonstrated that mice pretreated with the highly selective α 2-adrenoceptor antagonist CH-38083 showed blunting of the tumor necrosis factor-α (TNF-α) response induced by bacterial lipopolysaccharide (LPS). In the present study, the effect of a selective block of α 2-adrenoreceptor and the role of the sympathetic nervous system (SNS) in the regulation of LPS-induced TNF-α production was explored further using different selective adrenoceptor antagonists and agonists. While adrenalectomy did not prevent the effect of CH-38083, the block of the sympathetic transmission by chlorisondamine fully abolished the inhibitory effect of CH-38083 on LPS-induced TNF-α production, suggesting that the effect of the α 2-adrenoceptor blocking agent is corticosteroid-independent, but it requires intact sympathetic activity. Since the selective block of α 2-adrenoceptors results in an increased sympathetic activity and an increase of the release of noradrenaline (NA) in both the central and the peripheral nervous systems, and in our experiments propranolol, a non-selective β-adrenoceptor antagonist, and atenolol, a selective antagonist of β 1-adrenoceptors, prevented the effect of α 2-adrenoceptor blockade by CH-38083 of the TNF-α response induced by LPS, it seems likely that the excessive stimulation by NA of β 1-adrenoceptors is responsible for this action. The role of β-adrenoceptors and endogenous catecholamines is further substantiated by the finding that pretreatment of animals with propranolol alone resulted in a dose-dependent increase of the TNF-α response induced by LPS, and that isoproterenol, a non-selective β-adrenoceptor agonist, decreased it. Additionally, it was shown that prazosin, an α 1-and α 2 B -adrenoceptor antagonist, reduced LPS-induced TNF-α production. However, l-phenylephrine, a selective α 1-adrenoceptor agonist, was not able to modulate the TNF-α response following LPS challenge. Our findings that α 2-and β-adrenoceptor antagonists are able to decrease or increase, respectively, the TNF-α response elicited by LPS indicate that SNS, through release of endogenous catecholamines, is involved in vivo in the regulation of LPS-induced TNF-α production. In this process, the β-adrenoceptor-mediated events seems to play a pivotal role. Since the blockade of sympathetic activity by chlorisondamine failed to affect LPS-induced TNF-α release, it seems likely that, in vivo, the inhibitory effect of SNS on TNF-α production, mediated via β-adrenoceptors, is opposed by an effect of catecholamines on α 2-adrenoceptors. It is suggested that the fine-tuning of TNF-α release exerted by SNS in vivo might be particularly important during immunological and non-immunological stress, when the concentration of catecholamines is increased in the close proximity of TNF-α-secreting cells, which are known to possess adrenoceptors.