OBJECTIVES: We sought to assess prospectively whether patients with normal coronary angiograms but with impaired myocardial blood flow (MBF) increases to cold pressor testing (CPT) are at increased risk for cardiovascular events. BACKGROUND: Invasive angiographic assessments of coronary vasomotor function have demonstrated an impairment of endothelium-related coronary flow increases to independently predict future cardiovascular events. It is unknown whether noninvasive positron emission tomography (PET)-measured MBF alterations to sympathetic stimulation with CPT are associated with the risk of developing cardiovascular events. METHODS: A total of 72 patients (44 men, 28 women, age 58 ± 8 years) referred for diagnostic cardiac catheterization were studied. Myocardial blood flow was measured in absolute units with 13N- ammonia using PET, at baseline and during CPT in each patient. Cardiovascular events (cardiovascular death, acute coronary syndrome, myocardial infarction, percutaneous transluminal coronary angioplasty, coronary artery bypass grafting, ischemic stroke, or peripheral revascularization) were assessed as clinical outcome parameters over a mean follow-up period of 66 ± 8 months. Patients were assigned to three groups: group 1, patients with <40% increase in MBF (%ΔMBF), n = 22; group 2, patients with >0 and <40% increases in MBF, n = 32; and group 3, patients with decreases in MBF (≤0%), n = 18. RESULTS: During follow-up, one of the group 1 patients developed a cerebral stroke. In group 2, 15 cardiovascular events occurred in 9 patients and in group 3, 7 patients experienced 9 cardiovascular events (p ≤ 0.0001, univariate by log-rank test). Impaired MBF increases in group 2 and group 3 were associated with a significantly higher incidence of cardiovascular events by Kaplan-Meier analysis (p = 0.033, log-rank test). After adjusting for known coronary risk factors, MBF responses to CPT revealed a nonsignificant trend to be independently associated with a higher incidence for cardiovascular events (p = 0.065, multivariate by Cox regression model). CONCLUSIONS: Noninvasive PET-measured impaired MBF increases to sympathetic stimulation are associated with the risk of developing cardiovascular events.