Because of the complex physiology of the sympathetic nervous system, venous and urinary catecholamine levels in response to stressors may serve as approximations of sympathetic nervous system activity in the heart and vasculature. Furthermore, the sympathetic nervous system does not respond in a homogeneous fashion to stressors; instead, there is differential activation at various sites of the sympathetic nervous system. Despite these limitations, numerous studies document that peripheral venous samples reflect sympathetic nerve firing and are related to cardiovascular pathophysiology. Although arterial norepinephrine levels may be more sensitive than venous levels, venous levels do respond to stressors when attention is given to the time course of sampling, the imprecision of assays (and thus the requirement for larger sample sizes), and the selection of appropriate tasks. In addition to catecholamines, there are other compounds that may be used to index sympathetic nervous system responses to stressors. Chromogranin A, as measured in antecubital blood, responds to large-scale perturbations of the sympathetic nervous system but appears relatively insensitive to short-term behavioral challenge. In response to stressors, diastolic blood pressure is better predicted by plasma levels of renin than of norepinephrine.