The extent to which the sympathochromaffin system compared with other endocrine/neuroendocrine tissues contributes to the plasma chromogranin A pool has not been defined. To test the hypothesis that the sympathochromaffin system is the major source of circulating chromogranin A only when that system is activated markedly, we measured chromogranin A concentrations in 200 human plasma samples known to have a broad range of norepinephrine and epinephrine concentrations, reflecting therefore a broad range of sympathochromaffin activity at the time of sampling. Plasma chromogranin A and norepinephrine concentrations were highly correlated when the sympathochromaffin system was activated markedly (cardiac arrest samples, n = 13, r = 0.8392, P less than 0.0005) and when there was release of large amounts of norepinephrine from tumors (pheochromocytoma samples, n = 17, r = 0.8132, P less than 0.001). However, when the sympathochromaffin system was activated less markedly, resulting in plasma catecholamine concentrations that spanned the physiological and lower pathophysiological range (nonpheochromocytoma noncardiac arrest samples, n = 170), correlations between plasma chromogranin A and norepinephrine (r = 0.2877, P less than 0.0001) and epinephrine (r = 0.3814, P less than 0.0001) levels were relatively weak, although still statistically significant. Thus, at basal through moderate stress levels, norepinephrine and epinephrine concentrations accounted for only approximately 10-15% of the variance in plasma chromogranin A levels. We conclude that, although plasma chromogranin A concentrations are a valid marker of sympathochromaffin activity in humans, they are not a sensitive marker under physiological conditions.