Three immunometric assays of parathyroid hormone (PTH)--a commercial immunoradiometric assay, an in-house immunoradiometric assay, and an immunochemiluminometric assay--were evaluated in 50 patients with surgically proven primary hyperparathyroidism. Of these patients, 43 had increased values with the commercial assay (sensitivity, 86%), whereas 45 patients had increased concentrations with both the in-house immunoradiometric and the in-house immunochemiluminometric assays (sensitivities, 90%). Because of the results of this comparison study, we confidently chose the immunochemiluminometric assay as our routine assay; this assay was evaluated retrospectively in 361 patients with surgically proven primary hyperparathyroidism. In 45 patients, PTH values were below the upper limit of normal (sensitivity, 88%). The results indicate that the sensitivities of current immunometric assays are approximately 90%. Twenty patients who had hypercalcemia associated with malignant involvement were assessed with the immunochemiluminometric assay. Of these 20 patients, 19 had subnormal PTH values, and 1 had a value within the normal range. In contrast, in the past, PTH values determined with radioimmunoassays have often been in the normal range for such patients. Thus, an immunometric PTH assay is superior to a radioimmunoassay in the differential diagnosis of hypercalcemia associated with malignant disease.