OBJECTIVE: The pro-inflammatory cytokine tumor necrosis factor-alpha (TNF-alpha) impairs insulin action in insulin-sensitive tissues, such as fat, muscle and endothelium, and causes endothelial dysfunction. We hypothesized that TNF-alpha blockade with etanercept could reverse vascular and metabolic insulin resistance. METHOD AND RESULTS: Twenty obese patients with type 2 diabetes were randomized to etanercept treatment (25 mg subcutaneously twice weekly for 4 weeks) or used as controls in an open parallel study. Forearm blood flow and glucose uptake were measured during intra-arterial infusions of serotonin, sodium nitroprusside and insulin co-infused with serotonin. Beta-cell function was assessed with oral and intra-venous glucose tolerance tests and whole-body insulin sensitivity by hyperinsulinemic euglycemic clamps. Plasma levels of C-reactive protein and interleukin-6 decreased significantly with etanercept (C-reactive protein from 9.9 +/- 3.1 to 4.8 +/- 1.4 mg l(-1), p = 0.04; interleukin-6 from 3.1 +/- 0.4 to 1.9 +/- 0.2 ng l(-1), p = 0.03). Vasodilatory responses to serotonin and sodium nitroprusside infusions remained unchanged. Insulin effect on vasodilatation and on whole-body and forearm glucose uptake remained unchanged as well. Beta-cell function tended to improve. CONCLUSION: Although short-term etanercept treatment had a significant beneficial effect on systemic inflammatory markers, no improvement of vascular or metabolic insulin sensitivity was observed.