Objcetives:? Implantable cardioverter defibrillators (ICDs) are highly effective at preventing cardiac arrest, but their availability is limited by high cost. Automated external defibrillators (AEDs) are likely to be less effective, but also less expensive. We used decision analysis to evaluate the clinical and economic trade-offs of AEDs, ICDs, and emergency medical services equipped with defibrillators (EMS-D) for reducing cardiac arrest mortality. Methods:? A Markov model was developed to compare the cost-effectiveness of three strategies in adults meeting entry criteria for the MADIT II Trial: strategy 1, individuals experiencing cardiac arrest are treated by EMS-D; strategy 2, individuals experiencing cardiac arrest are treated with an in-home AED; and strategy 3, individuals receive a prophylactic ICD. The model was then used to quantify the aggregate societal benefit of these three strategies under the conditions of a constrained federal budget. Results:? Compared with EMS-D, in-home AEDs produced a gain of 0.05 quality-adjusted life-years (QALYs) at an incremental cost of $5225 ($104,500 per QALY), while ICDs produced a gain of 0.90 QALYs at a cost of $114,660 ($127,400 per QALY). For every $1 million spent on defibrillators, 1.7 additional QALYs are produced by purchasing AEDs (9.6 QALYs/$million) instead of ICDs (7.9 QALYs/$million). Results were most sensitive to defibrillator complication rates and effectiveness, defibrillator cost, and adults? risk of cardiac arrest. Conclusions:? Both AEDs and ICDs reduce cardiac arrest mortality, but AEDs are significantly less expensive and less effective. If financial constraints were to lead to rationing of defibrillators, it might be preferable to provide more people with a less effective and less expensive intervention (in-home AEDs) instead of providing fewer people with a more effective and more costly intervention (ICDs).