The use of direct oral anticoagulants for stroke prevention in patients with non-valvular atrial fibrillation (NVAF) is robust. However, the efficacy and safety of different dosage in patients with renal dysfunction is still a clinical challenge. We aimed to evaluate the clinical characteristics and outcomes of patients treated with apixaban in its different doses. A multicenter prospective cohort study, where consecutive eligible apixaban or warfarin treated patients with NVAF and renal impairment, were registered. Patients were followed-up for clinical events over a mean period of 1 year. Analyses were performed according to the dose of apixaban given, with consideration to the standard indications for dose reduction. Primary outcome was a composite of 1-year mortality, stroke or systemic embolism, major bleeding and myocardial infarction, while secondary outcomes included those components separated. Among the study population (n=2,140), risk of composite outcome was significantly lower in the high dose apixaban group (10%, n=491) than the low dose group (18%, n=673) and the warfarin group (18%, n=976) p <0.001. Results of 1-year mortality were similar. Apixaban dosing analysis revealed 65% of patients were appropriately dosed, while 31% were under-dosed and 4% were over-dosed. Furthermore, 53% of patients treated with low dose apixaban were under-dosed. Propensity score analysis revealed that patients who were appropriately treated with low-dose apixaban had a trend towards better composite outcome and mortality than 1:1 matched warfarin treated patients (18% vs 24%, p=0.09 and 16% vs 23%, p=0.06, respectively). Overall, appropriately dosed apixaban treated patients at any dose had significantly better outcomes than matched warfarin treated patients (composite outcome probability of 13.1% vs 18.6%, p=0.007). In conclusion, apixaban at any dose is a reasonable alternative to warfarin in patients with renal impairment, possibly associated with improved outcomes.