Subclinical vitamin D insufficiency is characterized by mild secondary hyperparathyroidism and enhanced risk of osteoporotic fracture. However, although low levels of 25-hydroxyvitamin D (25OHD) are common in otherwise normal elderly people, vitamin D status has not generally been taken into account in the previously published reference values for serum PTH. We measured fasting morning serum (obtained from April through June) PTH, total calcium, albumin, phosphate, creatinine, bone markers, and 25OHD in 280 healthy subjects (140 men and 140 women), aged 60-79 yr. Serum PTH was measured by means of 2 immunoradiometric assays, the Allegro intact PTH assay (Nichols Institute Diagnostics) and the new CAP assay (Scantibodies Laboratory, Inc.). We found a high prevalence (167 of 280; 59.6%) of low 25OHD (< or =30 nmol/L) in these otherwise healthy individuals. The PTH concentrations (95% confidence interval) obtained in the whole group of 280 subjects ranged from 13-64 ng/L for the Allegro assay and from 10-44 ng/L for the CAP assay. In the subjects with a serum 25OHD concentration greater than 30 nmol/L, values for both PTH assays were lower, 10-46 and 9-34 ng/L for the Allegro and the CAP assays, respectively. By using these values as a reference range, approximately 25% of the subjects with a serum 25OHD level of 30 nmol/L or less had a high serum PTH level (whatever the assay), reflecting secondary hyperparathyroidism. This might be missed if the reference PTH values are those obtained in the entire group, as is usually done. These results strongly suggest that vitamin D status should be taken into account when establishing reference values for serum PTH in elderly subjects.