Management of patients with atrial fibrillation (AF) is clinically challenging and has led to the development of new pharmacologic and nonpharmacologic therapies. However, no clear consensus on optimal endpoints for defining responses to therapy exists. This paradox arises largely because symptoms, often used to gauge efficacy of interventions, have not been well correlated with long-term "hard" endpoints such as stroke and death. One widely used symptom-based metric is "time to first symptomatic AF episode," but this correlates poorly with frequency of symptomatic episodes. Similarly, although quality-of-life (QOL) measures have been used, the precise and unbiased assessment of QOL is difficult to define and measure. The availability of implantable devices capable of monitoring and recording all AF episodes has made the accurate determination of total time in AF ("burden") possible. However, QOL tools and formal measures of AF burden do not correlate well, suggesting that measures of subjective well-being are important adjunct measures to conventional measures of disease severity when evaluating the therapeutic efficacy of treatments for AF. Although measurement of total AF burden requires invasive monitoring, symptomatic AF burden (defined by frequency, duration and severity of symptoms) can be determined in all patients with symptomatic AF and may serve as a valid endpoint, as elimination of symptoms is a common and realistic therapeutic goal. Therefore, we propose an algorithm to quantify symptomatic AF burden as an endpoint in clinical trials.