To assess the incidence of atrial fibrillation (AF) and the clinical impact of AF types on outcomes in patients with acute myocardial infarction (AMI) treated invasively, we analyzed 2,980 consecutive patients with AMI admitted to our department from 2003 through 2008. Data collected by the insurer were screened to identify patients who died during the median follow-up of 41 months. AF was recognized in 282 patients (9.46%, AF group); the remaining 2,698 patients (90.54%) were free of this arrhythmia (control group). The AF group was divided into 3 subgroups: prehospital paroxysmal AF (n = 92, 3.09%), new-onset AF (n = 109, 3.66%), and permanent AF (n = 81, 2.72%). In-hospital and long-term mortalities were significantly higher (p0.001 for the 2 comparisons) in the AF than in the control group (14.9% vs 5.3%, 37.2% vs 17.0%, respectively). Long-term mortality was significantly higher (p0.001 for the 2 comparisons) in the new-onset AF (35.8%) and permanent AF (54.3%) groups than in the control group but did not differ significantly between the prehospital AF and control groups (21.7% vs 17.0%, p = NS). Considering types of arrhythmia separately, only permanent AF (hazard ratio 2.59) was an independent risk factor for death in the studied population. In conclusion, AF occurs in 1 of 10 patients with AMI treated invasively, with nearly equal distributions among prehospital, new-onset, and permanent forms. Although arrhythmia is a marker of worse short- and long-term outcomes, only permanent AF is an independent predictor for death in this population.