Pigmented and nonpigmented lesions in 49 patients were studied to determine the reliability of a test based on the concentration of intravenously injected radioactive phosphorus (P32) in such lesions. Counts were made frequently over a three-hour period at the site of the tumor and were compared with counts made at a contralateral normal site in the same patient. Two illustrative cases are described. In the first case the result of the test was positive and the lesion when excised proved to be a melanoma. In the second case, however, results of the test suggested a melanoma but the excised tumor proved to be a benign sclerosing hemangioma. In the entire series there were 20% false-positive and 37% false-negative results. In its present form this test cannot be relied on to distinguish melanomas from other skin tumors.