The new anticoagulants dabigatran and rivaroxaban can be responsible for haemorrhagic complications. As for any anticoagulant, bleeding management is challenging. We aimed to test the effect of all putative haemostatic agents on the anticoagulant activity of these new drugs using thrombin generation tests. In an ex vivo study, 10 healthy white male subjects were randomised to receive rivaroxaban (20 mg) or dabigatran (150 mg) in one oral administration. After a two weeks washout period, they received the other anticoagulant. Venous blood samples were collected just before drug administration (H0) and 2 hours thereafter. Reversal of anticoagulation was tested in vitro using prothrombin complex concentrate (PCC), rFVIIa or FEIBA® at various concentrations. Rivaroxaban affects quantitative and kinetic parameters, including the endogenous thrombin potential (ETP-AUC and more pronouncedly the thrombin peak), the lag-time and time to peak. PCC strongly corrected ETP-AUC, whereas rFVIIa only modified the kinetic parameters. FEIBA corrected all parameters. Dabigatran specially affects the kinetics of thrombin generation with prolonged lag-time and time to peak. Although PCC increased ETP-AUC, only rFVIIa and FEIBA corrected the altered lag-time. For both anticoagulants, lower doses of FEIBA, corresponding to a quarter to half the dose usually used, have potential reversal profile of interest. In conclusion, some non-specific reversal agents appear to be able to reverse the anticoagulant activity of rivaroxaban or dabigatran. However, clinical evaluation is needed regarding haemorrhagic situations, and a meticulous risk-benefit evaluation regarding their use in this context is required.