OBJECTIVES: This real-world study evaluated the baseline characteristics of patients with non-valvular atrial fibrillation (NVAF) who had switched from warfarin to novel oral anticoagulants (NOACs). METHODS: Retrospective cohort study was conducted using the MarketScan^® plus Earlyview data from 10/1/2009 to 12/31/2013.  Adult NVAF patients (ICD-9 code 427.31 or 472.32) with one year of baseline period and a history of continuous warfarin use in the baseline period for at least 3 months immediately before the index date (defined as the first NOAC claim) were included.  Patients with evidence of valvular heart disease, thyrotoxicosis, pericarditis, mitral stenosis, VTE, cardiac surgery, and endocarditis during the baseline period were excluded.  Categorical variables were reported as percentages and frequencies, and continuous variables as means±SD.  Categorical variables were compared using Pearson's chi-squared test while continuous variables were compared using wilcoxon signed-rank test. RESULTS: Among 11,743 eligible patients, 427 (3.64%) switched to apixaban, 8,989 (76.55%) to dabigatran and 2,327 (19.81%) to rivaroxaban.  Apixaban (74.82±11.39 years) patients were older versus those who switched to dabigatran (72.46±10.89 years, p<0.0001) or rivaroxaban (73.50±11.27 years, p=0.0193).  Apixaban users (45.4%) were more female compared to dabigatran (38.4%, p=0.0037) and rivaroxaban (40.4%, p=0.0499).  The mean CHADS score was higher for apixaban users (mean±SD 2.28±1.25) as compared to dabigatran (1.94±1.20, p<0.0001) and rivaroxaban (2.18±1.25, p=NS) users.  Apixaban patients had significantly higher baseline rates for congestive heart failure (p=0.0111), hypertension (p=0.0002), renal disease (p=0.0017) and ischemic stroke/ transient ischemic attack (p=0.0004) as compared to dabigatran users.  Apixaban users (2.34±2.12) also had higher mean charlson comorbidity index scores as compared to dabigatran users (1.98±1.96, p=0.0002). CONCLUSIONS: Patients who switch to apixaban are older and sicker as compared to those switching to dabigatran or rivaroxaban.  A detailed evaluation of patient characteristics on the treatment outcomes in NVAF patients switching from warfarin to NOAC is warranted in future.