Vitamin D requirements are thought to vary with age, but there is little comparative evidence for this. One goal in establishing a vitamin D requirement is to avoid secondary hyperparathyroidism. We studied 1741 euthyroid, thyroid clinic outpatients without evidence of calcium abnormalities, ranging in age from 19 to 97 yr, whose serum and urine had been analyzed for calcium, vitamin D, and parathyroid status. We found no effect of age on the 25-hydroxyvitamin D [25(OH)D] concentration associated with specific vitamin D intakes, and there was no relationship between 25(OH)D and 1,25hydroxyvitamin D [1,25(OH)2D]. In every age group, serum 1,25(OH)2D declined with increasing creatinine (P < 0.001). What changed with age included creatinine, which correlated with 25(OH)D (r = 0.146, P < 0.001) only in the youngest age group (19-50 yr) but not in the older age groups (P > 0.1). Creatinine did not correlate with PTH in the youngest age group, but the relationship became significant as age increased (e.g. for the elderly, r = 0.365, P < 0.001). Linear regression of log PTH vs. log 25(OH)D agreed with the natural shape of the relationship observed with scatterplot smoothing, and this showed no plateau in PTH as 25(OH)D increased. We compared PTH concentrations among age groups, based on 20 nmol/liter increments in 25(OH)D. Mean PTH in adults older than 70 yr was consistently higher than in adults younger than 50 yr (P < 0.05 by ANOVA and Dunnett's t test). PTH levels of the elderly who had 25(OH)D concentrations greater than 100 nmol/liter matched PTH of younger adults having 25(OH)D concentrations near 70 nmol/liter. This study shows that all age groups exhibit a high prevalence of 25(OH)D insufficiency and secondary hyperparathyroidism. Older adults are just as efficient in maintaining 25(OH)D, but they need more vitamin D to produce the higher 25(OH)D concentrations required to overcome the hyperparathyroidism associated with their diminishing renal function.