The decision to operate on a patient with an aortic aneurysm is based on the risk of aneurysm rupture versus the risk of aneurysm repair, within the context of the patient's overall life expectancy. Risk of rupture is still primarily based on the maximum aneurysm diameter, with some allowances made for factors that modify rupture risk, such as gender and current smoking. Newer methods for determining rupture risk, such as aneurysm-wall stress analysis, appear promising, but are not yet broadly available. Until then, diameter-based prediction rules for rupture risk will "fail" 10% to 25% of patients with both small and large abdominal aortic aneurysms. With regard to predicting operative mortality and life expectancy after open or endovascular aneurysm repair, multiple risk-stratification algorithms have been created. The best of these algorithms are accurate in 75% to 80% of patients, meaning that they fail in 20% to 25% of cases. Prediction algorithms provide significant guidance, but cannot take the place of an experienced clinician at this point. Somehow, experienced surgeons are able to sift through a massive amount of information and properly select patients who are appropriate for surgery, with quite reasonable perioperative and long-term mortality rates.