Whether trough-phase rivaroxaban concentrations provide sufficient anticoagulation needs more study. We evaluated levels of coagulation activation markers in the trough concentration phase in nonvalvular atrial fibrillation (NVAF) patients, and the correlation between these markers and rivaroxaban concentration. Fifty-five Japanese NVAF patients received 24-week rivaroxaban treatment of either 15 or 10 mg once-daily in the morning. Of these, 26 patients had no history of anticoagulant therapy (naive group) and 29 had switched from warfarin (warfarin group). D-dimer and prothrombin fragment 1 + 2 (F1 + 2) levels, and protein C activities were measured at 0 (baseline), 12 and 24 weeks of rivaroxaban treatment just before the patient's regular dosing time (trough phase). For 49 patients, D-dimer, F1 + 2, and rivaroxaban concentrations were also measured twice between 28 and 32 weeks of rivaroxaban treatment at non-trough times to achieve a range of drug concentrations for correlation analysis. For the naive group, D-dimer and F1 + 2 levels were significantly reduced (p < 0.01) from baseline at 12 and 24 weeks. For the warfarin group, these values were unchanged for D-dimer but significantly increased (p < 0.01) for F1 + 2. Protein C activity was unchanged in the naive group and was increased (p < 0.01) in the warfarin group. Prothrombin time (r = 0.92, p < 0.0001) and activated partial thromboplastin time (r = 0.54, p < 0.0001) correlated with rivaroxaban concentration, but not D-dimer and F1 + 2 levels. In conclusion, rivaroxaban in the trough phase is comparable to warfarin in reducing D-dimer levels. Although trough level rivaroxaban suppresses F1 + 2 less than warfarin, the higher activities of protein C with rivaroxaban treatment compared to warfarin treatment may counterbalance this. Lack of correlation between rivaroxaban concentration and D-dimer and F1 + 2 levels suggests that trough concentrations of rivaroxaban reduce their concentrations as effectively as higher levels do.