Abstract Objective: No studies have been conducted in Turkey to compare the quality of stroke prevention therapies provided in different healthcare settings in patients with atrial fibrillation (AF). Therefore, we aimed to evaluate possible differences between secondary (SH) and tertiary hospital (TH) settings in the effectiveness of implementing AF treatment strategies. Methods: Baseline characteristics of 6273 patients with non-valvular AF enrolled in the RAMSES (ReAl-life Multicentre Survey Evaluating Stroke Prevention Strategies in Turkey) study were compared. Results: Of the study population, 3312 (52.8%) patients were treated in THs and 2961 (47.2%) patients were treated in SHs. Patients treated in the SH setting were older (70.8±9.8 vs. 68.7±11.4years, p<0.001), had a lower socioeconomic status, had a higher CHA2DS2VASc and HASBLED scores (3.4±1.4 vs. 3.1±1.7, p<0.001 and 1.7±1.0 vs. 1.6±1.1, p<0.001 respectively), and had more comorbidities than patients treated in THs. Inappropriate oral anticoagulant use was more prevalent in SHs than THs (31.4% vs. 25.6%, p<0.001). When over- and undertreatment rates were compared among hospital types, overtreatment was more prevalent in THs (7.6% vs. 0.9%, p<0.001) while undertreatment was more common in SHs (30.5% vs. 17.9%, p<0.001). Conclusion: This study demonstrates the marked disparity between patient groups with AF presenting at SHs and THs. The use of guideline-recommended therapy is not adequate in either type of centre, overtreatment was more prevalent in THs and undertreatment was more prevalent in SHs.