The clinical utility of the urinary cyclic AMP:creatinine ratio in assessing parathyroid function was evaluated in 33 hyper-calcemic patients and compared this with the determination of the renal component of urinary cyclic AMP. We found the discriminatory value of urinary cyclic AMP:creatinine ratio to be slightly superior and to have additional advantages. Not only did the urinary cyclic AMP:creatinine ratio show empirically somewhat better discrimination between normals and patients with primary hyperparathyroidism, but it is technically simpler than the determination of the nephrogenous cyclic AMP. Our urinary cyclic AMP excretion data show 90% discrimination of primary hyperparathyroid subjects from normals. Among all hypercalcemic patients studied who had both elevated urinary cyclic AMP and elevated parathyroid hormone (PTH) levels by radioimmunoassay (RIA), 77% had primary hyperparathyroidism, and 23% had malignancy-associated hypercal-cemia. Of those patients with malignant tumors and hypercalcemia, half had elevated urinary cyclic AMP and two thirds had elevated PTH by RIA. These data suggest that these tests have little discriminatory value in differentiating primary hyperparathyroidism from malignancy-associated hypercalcemia. No hypercalcemic patient who had both serum PTH and urine cyclic AMP in the normal range was found to have primary hyperparathyroidism. This suggests that further observation and evaluation is indicated in such patients before exploratory surgery is undertaken.