OBJECTIVE Subarachnoid hemorrhage (SAH) frequently results in myocardial necrosis with release of cardiac enzymes. Historically, this necrosis has been attributed to coronary artery disease, coronary vasospasm, or oxygen supply-demand mismatch. Experimental evidence, however, indicates that excessive release of norepinephrine from the myocardial sympathetic nerves is the most likely cause. We hypothesized that myocardial necrosis after SAH is a neurally mediated process that is dependent on the severity of neurological injury. METHODS Consecutive patients admitted with SAH were enrolled prospectively. Predictor variables reflecting demographic (age, sex, body surface area), hemodynamic (heart rate, systolic blood pressure), treatment (phenylephrine dose), and neurological (Hunt-Hess score) factors were recorded. Serial cardiac troponin I measurements and echocardiography were performed on days 1, 3, and 6 after enrollment. Troponin level was treated as a dichotomous outcome variable. We performed univariate and multivariate analyses on the relationships between the predictor variables and troponin level. RESULTS The study included 223 patients with an average age of 54 years. Twenty percent of the subjects had troponin I levels >1.0 microg/L (range, 0.3 to 50 microg/L). By multivariate logistic regression, a Hunt-Hess score >2, female sex, larger body surface area and left ventricular mass, lower systolic blood pressure, and higher heart rate and phenylephrine dose were independent predictors of troponin elevation. CONCLUSIONS The degree of neurological injury as measured by the Hunt-Hess grade is a strong, independent predictor of myocardial necrosis after SAH. This finding supports the hypothesis that cardiac injury after SAH is a neurally mediated process.