Atrial fibrillation (AF) is the most common arrhythmia seen in clinical practice and results in the largest number of arrhythmia related hospital admissions. Despite the enormity of its impact on patients and the health care system, current medical therapy for AF is inadequate. Therapeutic approaches have been guided by understanding of fibrillation mechanisms. AF results from multiple simultaneous reentrant wavefronts. This is a diffuse and dynamic substrate with no discrete anatomic target. The catheter mediated "Maze" procedure employs creation of linear lesions to divide the atria into segments too small to support reentrant activation. Using current catheter technology this is a challenging procedure with low success rates and high complication rates. The observation that rapid focal firing of atrial myocytes within the pulmonary veins initiates fibrillation in the majority of paroxysmal AF patients has led to an entirely new ablation strategy. The sites of firing that initiate AF are targeted for ablation. Thus the paradigm for AF ablation has changed dramatically from altering the substrate of ongoing fibrillation to elimination of the triggers that initiate fibrillation. Initial experience revealed that multiple sites in the pulmonary veins are capable of rapid firing. Unfortunately not all sites fire during an ablation procedure. Sites that are quiescent during an ablation procedure may result in AF recurrence despite acute success. Ablation strategy has thus changed yet again to electrical isolation of all pulmonary venous tissue from the left atrium. The evolution of ablation strategies has paralleled our understanding of AF mechanisms. Elucidation of the mechanisms responsible for venous firing may lead to more specific therapy for the prevention of AF in the future.