Urinary cyclic AMP (UcAMP) appropriate for the serum calcium concentration was determined in normal subjects during the base-line state and during alteration in their serum calcium concentrations by saline and calcium infusions. This was compared to the UcAMP in 76 patients with hypercalcemia and 5 patients with hypocalcemia. In 54 of 56 patients with primary hyperparathyroidism, the UcAMP was inappropriately high for their serum calcium concentration, the 2 exceptions having renal failure. In four patients with vitamin D intoxication, sarcoidosis, milkalkali syndrome, and thiazide-induced hypercalcemia and in five patients with hypocalcemia due to hypoparathyroidism, the UcAMP was appropriately low for their serum calcium concentration. In 16 patients with nonparathyroid neoplasms, 10 had UcAMP levels that were inappropriately high suggesting ectopic parathyroid hormone (PTH)-mediated hypercalcemia and 6 had UcAMP levels that were appropriately low suggesting that their hypercalcemia was due to osteolytic factors other than PTH. Correlations between UcAMP, serum calcium concentration, and carboxyl-terminal immunoreactive PTH suggest that random UcAMP is a sensitive accurate reflection of circulating biologically active PTH. If there is adequate renal function (serum creatinine concentration less than 2.0 mg/dl), a random UcAMP expressed as mumol/g creatinine and analyzed as a function of the serum calcium concentration completely separates patients with PTH and non-PTH-mediated hypercalcemia.