Background The dilute Russell's viper venom time (dRVVT) has been suggested for the assessment of the intensity of anticoagulation of all direct oral anticoagulants (DOACs). This study aimed to compare the performance of an optimized liquid-stable dRVVT-based DOAC assay (DRVV-DOAC) on clinical samples before and after mixing these with normal pooled plasma (NPP). Methods Forty-one apixaban, 25 dabigatran, 56 rivaroxaban and 49 vitamin K antagonists (VKAs) plasma samples were included for retrospective analysis. Plasma DOAC concentrations were determined by liquid chromatography coupled with tandem mass-spectrometry. INR was determined for all VKA samples. DRVV-DOAC was performed with an original ready-to-use reagent (Haematex Research ™) where plasma samples were tested neat and in a 1:1 mix with NPP. Results Plasma concentrations ranged from 1 to 406 ng/ml for apixaban, 0 to 386 ng/ml for dabigatran and 0 to 719 ng/ml for rivaroxaban. INR ranged from 2.2 to 6.1. DRVV-DOAC correlated well with plasma concentrations (r 2 = 0.70, 0.94, 0.63 (non-mixed procedure) and 0.77, 0.97, 0.86 (mixed procedure) for apixaban, dabigatran and rivaroxaban, respectively). DRVV-DOAC measurements in the normal range ruled out dabigatran and rivaroxaban concentrations above 30 and 50 ng/ml, but performance was lower for apixaban. DRVV-DOAC was sensitive to VKA samples but poorly reflected INR values. When VKA samples were mixed with NPP, DRVV-DOAC measurements decreased to values close to baseline clotting time. Conclusions DRVV-DOAC is a quick method which showed increased sensitivity compared with other phospholipid-rich dRVVT reagents already investigated. Mixing samples with NPP improved the specificity but reduced sensitivity, especially for apixaban.