Complement plays an essential role in inflammation and tissue damage. However, it is largely unknown to what extent the system acts as a primary inducer of secondary mediator systems in the inflammatory network of human whole blood. Here we describe a novel in vitro model using the thrombin-specific hirudin analog lepirudin as anticoagulant, which, in contrast to heparin, did not interfere with complement activation. The model was used to study the role of complement in Escherichia coli-induced inflammatory responses. Granulocyte and monocyte oxidative burst was complement dependent as it was reduced by 85% and 70%, respectively, by the C3 [corrected] binding peptide compstatin. A similar reduction was found by inhibition of C5, C5a, and C5a receptor (C5aR). Furthermore, anti-CR3 antibodies were as efficient as the C5aR antagonist in reducing granulocyte oxidative burst, whereas blocking CD14 or C3aR had no effect. Up-regulation of granulocyte CR3 was virtually abolished by a C5aR antagonist. Opsonization and phagocytosis was completely inhibited by blocking of C5aR or CR3, whereas blocking of the FcgammaRs (CD16, CD32, CD64) had no effect. In contrast to oxidative burst and phagocytosis, cytokine secretion was largely complement independent. Thus, anti-CD14 abolished tumor necrosis factor-alpha, interleukin-6 (IL-6), and IL-10 secretion, whereas IL-8 was equally inhibited by anti-CD14 and compstatin. In conclusion, the present model is particularly useful for studying complement as part of the inflammatory network. The results emphasize a crucial role for C5a-C5aR interaction in E coli-induced up-regulation of CR3 and the subsequent oxidative burst and phagocytosis. Complement inhibition may have therapeutic implications in oxidative burst-induced tissue damage.