Measurements of plasma normetanephrine and metanephrine provide a highly sensitive test for diagnosis of pheochromocytoma, but false-positive results remain a problem. We therefore assessed medication-associated false-positive results and use of supplementary tests, including plasma normetanephrine responses to clonidine, to distinguish true- from false-positive results. The study included 208 patients with pheochromocytoma and 648 patients in whom pheochromocytoma was excluded. Clonidine-suppression tests were carried out in 48 patients with and 49 patients without the tumor. Tricyclic antidepressants and phenoxybenzamine accounted for 41% of false-positive elevations of plasma normetanephrine and 44-45% those of plasma and urinary norepinephrine. High plasma normetanephrine to norepinephrine or metanephrine to epinephrine ratios were strongly predictive of pheochromocytoma. Lack of decrease and elevated plasma levels of norepinephrine or normetanephrine after clonidine also confirmed pheochromocytoma with high specificity. However, 16 of 48 patients with pheochromocytoma had normal levels or decreases of norepinephrine after clonidine. In contrast, plasma normetanephrine remained elevated in all but 2 patients, indicating more reliable diagnosis using normetanephrine than norepinephrine responses to clonidine. Thus, in patients with suspected pheochromocytoma and positive biochemical results, false-positive elevations due to medications should first be eliminated. Patterns of biochemical test results and responses of plasma normetanephrine to clonidine can then help distinguish true- from false-positive results.