BACKGROUND AND OBJECTIVE To determine the adverse effects (AE) of antiarrhythmic drugs used to treat atrial fibrillation (AF) in routine clinical practice. PATIENTS AND METHODS A retrospective, observational, population-based study of all patients diagnosed of AF requiring long-term treatment with antiarrhythmic drugs. We performed also a descriptive analysis of the type of EA, and their management (diagnostic techniques, laboratory tests, drug treatment, number of visits to health services, emergency visits, hospitalizations and surgeries). RESULTS We included 400 patients. Overall, 688 treatments were started, 34 of which (4.9%) were implied in 32 EA in 30 patients. The most frequent EA were hypothyroidism (28.1%), hyperthyroidism (25%) and bradycardia (9.38%). In absolute numbers, class III drugs (Vaughan-Williams classification) were associated with a higher number of AE (p=0.001). The number of AE/100 patients-year was: class I 0.42 (0.01-1.15%), class II 2.55 (0.68-5.13%), class III 4.70 (2.98-6.64%), class IV 1.18 (0.14-2.80%), digoxin 0.41 (0.08-1.19). Individually, the drugs with a higher rate were sotalol, bisoprolol and amiodarone (5.70, 5.37 and 4.59 respectively). 93.75% of EA required visits to health services, 59.38% laboratory tests and 56.25% diagnostic techniques. The most frequent visits were scheduled for primary care (75%), followed by scheduled visits to outpatient cardiology (21.9%). 71.9% of EA were resolved without any sequel and one patient died. CONCLUSIONS Antiarrhythmic drugs for the management of AF are associated with the presence of AE, mainly class III drugs. In most cases they are not serious, but imply important health resources.