Recurrent implantation failure (RIF), defined as the absence of clinical pregnancy after the transfer of three good-quality embryos, concerns up to 40% of IVF couples and is associated with a low success rate. The causes remain unexplained in over 50% of cases. Various dysimmune changes (related to immune T cells profiles, pro-inflammatory cytokines levels) have been described in unexplained RIF as compared to fertile controls, and it has been estimated that such dysimmunity may occur in 50% of unexplained RIFs. Previous data on a benefit of general immune modulation by steroids or immunoglobulins are heterogenous and failed to demonstrate clinically significant benefit. The proinflammatory cytokine Tumor Necrosis Factor (TNF) α participates in the regulation of the immune balance of the endometrium, its peripheral blood and endometrial concentrations are increased in RIF patients as compared to fertile controls. In 2009, a pilot placebo controlled study showed that TNF-α antagonist treatment allowed a 56% live birth rate (versus 13% in controls) in 13 women with unexplained RIF. Due to the lack of maternal and fetal tolerance data, TNF-α antagonists were not further evaluated. Today, safety data issued from 1200 pregnancies are reassuring allowing the use of TNF-α antagonists during pregnancy (www.lecrat.org). In addition the TNF-α antagonist certolizumab does not cross the placental barrier. We hypothesize that certolizumab may improve clinical pregnancy rates in women with unexplained RIF with a good safety profile.