OBJECTIVES: To assess convergent and divergent validity of a new questionnaire (SAFUCA) developed to measure satisfaction with medical care in patients with non valvular atrial fibrillation. METHODS: The 25-item reduced version instrument was administered along with the Treatment Satisfaction Questionnaire for Medication (TSQM), and 5 Visual Analog Scales (VAS) measuring: Health Related Quality of Life (HRQoL), Effectiveness and Overall Satisfaction assessed by the patient; and Effectiveness and Tolerability assessed by the clinician. A convenience sample of 230 patients recruited at 7 health centers (5 hospitals and 2 primary care centers) in Spain was used. Second Order Exploratory Factor Analysis (SO-EFA) and correlation with VAS scales were computed. RESULTS: SAFUCA dimensions correlated (p<0.001) higher with corresponding TSQM dimensions (e.g.: Effectiveness r=0.450, Convenience r=0.457, Undesired Effects r=-0.340, Overall satisfaction r=0.651), while SO-EFA made evident differences regarding the assessment of satisfaction with INR controls, interference in QoL and Medical Care. Correlation pattern between VAS scores was significantly different between primary care centers and hospitals, with a higher correlation between patient and clinician scores observed in primary care (e.g.: Effectiveness rPC=0.430 vs. rH=0.057). Similarly, correlation patterns between VAS concurrent scales and SAFUCA dimensions differed between center types. While high correlations between Effectiveness dimension and patient satisfaction VAS were observed in both cases (rPC=0.521 vs. rH=0.516), correlation of Tolerability with Undesired Effects differed between centers (rPC=-0.474 vs. rH=-0.085). CONCLUSIONS: The 25-item questionnaire exhibits good convergent and divergent validity values. Differences between types of health care centers in correlation patterns were meaningful and worth of further research.