The inhibition of tumor necrosis factor (TNF) through the use of either antibodies or soluble receptors is a highly effective strategy for the clinical control of chronic inflammatory conditions such as rheumatoid arthritis. Different viruses have similarly exploited this concept by expressing a set of specifically tailored secreted TNF decoy receptors to block host inflammatory responses. Poxviruses have been shown to encode at least two distinct molecules, termed Cytokine response modifier D (CrmD) and CrmB, in which a TNF inhibitor is combined with a chemokine inhibitor on the same molecule. The ectromelia virus CrmD protein was found to be a critical determinant of virulence in vivo, being able to control local inflammation to allow further viral spread and the establishment of a lethal infection. Strikingly, both the TNF and the chemokine inhibitory domains are required for the full activity of CrmD, suggesting a model in which inhibition of TNF is supported by the concomitant blockade of a reduced set of chemokines. Inspired by this model, we reasoned that a similar strategy could be applied to modify the clinically used human TNF receptor (etanercept), producing a generation of novel, more effective therapeutic agents. Here we show the analysis of a set of fusion proteins derived from etanercept by addition of a viral chemokine-binding protein. A bifunctional inhibitor capable of binding to and blocking the activity of TNF as well as a set of chemokines is generated that is active in the prevention of arthritis in a murine disease model.