During the past 10 years numerous studies on the treatment of paroxysmal atrial fibrillation (AF) by right and left atrial ablation procedures have been published. The results of studies based on follow-up periods of a few months have been repeatedly interpreted as providing evidence for curative therapy. However, insufficient focus on the variability of the natural history of paroxysmal AF, the inadequate detection of silent arrhythmic events, the eclectic post-interventional use of antiarrhythmic drugs, and the lack of appropriate control groups make the reports unconvincing. Randomized controlled trials are needed to confirm postulated long-term cure rates for AF.