Abstract Background The lifetime risk of developing Atrial Fibrillation (AF) is one in three adults, with a prevalence ranging from 2% to 4%. In 2017, the Evaluated Heart valves, Rheumatic or Artificial (EHRA) classification was introduced to guide oral anticoagulation (OAC) for patients with AF and valvular heart disease. Objective To analyze the evolution of patients with AF according to the EHRA categorization type 1, type 2, and those without valvular heart disease (group 3). Method Prospective, observational, and longitudinal study with 421 patients with AF, divided into three groups. Baseline risk scores for embolism, bleeding, and OAC quality were calculated. Events were verified during the 12-month clinical follow-up. Chi-square, parametric, and non-parametric tests were used for statistical analysis, in addition to the Kaplan-Meier curve. P-value < 0.05 was the statistical significance criterion. Results The average age was 58.6 years; 227 patients were women and 269 had OAC. There were 113 OAC EHRA type 1 patients, 53 type 2, and 255 patients from group 3. Age was lower, the proportion of women and OAC usage was higher in group 1 and there was a lower ejection fraction in group 3. During follow-up, the OAC percentage was 87.6% in type 1; 62.2% in type 2; and 56.0% in group 3 (p < 0.0001), with no difference regarding OAC quality, bleeding rate, embolism, or combined events. A higher total mortality rate was observed in group 3 (p = 0.02). Conclusion Despite the higher use of OAC in EHRA type 1, there were no differences between the groups regarding bleeding, embolism, or combined events. Group 3, characterized by greater systolic dysfunction, exhibited higher total mortality rates. (AF adults 2 4 4% 2017 valves (EHRA (OAC . 3) Prospective observational 42 embolism calculated 12month month 12 followup. followup follow up. up follow-up Chisquare, Chisquare Chi square, square Chi-square parametric nonparametric non analysis KaplanMeier Kaplan Meier curve Pvalue P value 005 0 05 0.0 criterion 586 58 6 58. years 22 26 11 5 25 followup, up, 876 87 87.6 622 62 62.2 560 56 56.0 p 0.0001, 00001 0.0001 , 0001 0.0001) events 0.02. 002 0.02 02 0.02) dysfunction rates 201 00 0. 8 87. 62. 56. 0000 0.000 000 20 0.00