The new direct-acting anticoagulants such as dabigatran and rivaroxaban are usually not monitored but may be associated with haemorrhage, particularly where renal impairment occurs. They have no effective "antidotes". We studied 17 patients receiving dabigatran 150 mg twice daily for non-valvular atrial fibrillation and 15 patients receiving rivaroxaban 10 mg daily for the prevention of deep venous thrombosis after hip or knee replacement surgery. We assessed the effect of these drugs on commonly used laboratory tests and Calibrated Automated Thrombogram (CAT) using plasma samples. We also assessed effects in fresh whole blood citrated patient samples using thromboelastography on the TEG and the ROTEM. The efficacy of nonspecific haemostatic agents prothrombin complex concentrate (PCC), Factor VIII Inhibitor By-passing Activity (FEIBA) and recombinant activated factor VII (rVIIa) were tested by reversal of abnormal thrombin generation using the CAT. Concentrations added ex vivo were chosen to reflect doses normally given in vivo. Dabigatran significantly increased the dynamic parameters of the TEG and ROTEM and the lag time of the CAT. It significantly reduced the endogenous thrombin potential (ETP) and reduced the peak height of the CAT. Rivaroxaban did not affect the TEG and ROTEM parameters but did increase the lag time and reduce ETP and peak height of the CAT. For both drugs, these parameters were significantly and meaningfully corrected by PCC and FEIBA and to a lesser but still significant extent by rFVIIa. These results may be useful in devising a reversal strategy in patients but clinical experience will be needed to verify them.