We have found recently that excluding subjects with low serum 25OHD has a significant impact on the PTH reference range (10-46 ng/liter instead of 10-65 ng/liter with the same assay). However, before being used routinely, this new range had to be clinically validated. We thus reviewed the chart of 708 consecutive osteopenic patients who were referred to our unit for a biological exploration in search of secondary causes for their low bone mass. They were classified into two groups. Group 1 (n = 360) included the patients for whom no reasons for high PTH were found after examination of their chart. Group 2 (n = 348) included patients with one of the following potential reasons for an increased PTH concentration: hyper- or hypocalcemia, normocalcemic primary hyperparathyroidism (PHPT), renal hypercalciuria, vitamin D insufficiency, chronic renal failure, use of bisphosphonates, and any chronic disease known to potentially alter calcium metabolism. Among the 360 group 1 patients, 15 (4.2%) had a serum PTH level more than 46 ng/liter, which is not different from the theoretical rate of 3% of normal subjects whose serum PTH may be above the 97th centile of the reference (chi(2) = 2.8; NS). Forty-two group 2 patients had a surgically proven PHPT. Among these, serum PTH was </=65 ng/liter in 17 (40.5%) and </=46 ng/liter in 5 (12%). In conclusion, our proposed PTH reference range allows to identify fewer patients with mild surgically proven PHPT who have a normal serum PTH concentration, without inducing an increase in the rate of falsely high PTH.