We studied the effects of increasing age, dosage, and duration of parathyroid hormone (PTH) treatment on changes in bone mineral density (BMD). Randomized placebo controlled trials on PTH treatment in men or women were retrieved from PubMed (1951 to present), Web of Science (1945 to present), or Embase (1974 to present). The search date was November 16, 2010. All studies comparing PTH treatment to either placebo or antiresorptive drugs--for example, bisphosphonates or hormone replacement therapy--were included. A total of 214 studies were identified in the initial search, and 15 of these trials were included. By metaregression analysis, we found that the increase in spine BMD (Z-score) after PTH treatment was blunted by increasing age (R(2) = 0.27; 2p = 0.01, slope -0.023 Z-scores per year, 11 studies). By increasing PTH dosage (μg/d), spine BMD increased significantly (2p = 0.002) with a slope of +0.011 Z-scores/μg/d of teriparatide. Furthermore, the duration of treatment was positively correlated to spine BMD (P < 0.001) with a slope of +0.043 Z-score for each extra month of treatment. We evaluated the BMD effect in hips and found no age dependency (R(2) = 0.04; P = 0.66; 8 studies). However, for the spine, we found a significant relation to daily dosage (P = 0.011), Z-score coefficient 0.0051 ± 0.0020 (2p < 0.01). The treatment duration also correlated positively by a Z-score coefficient of 0.0170 ± 0.0053, 2p < 0.01 per extra month of treatment. PTH treatment alone seems to be able to improve BMD significantly. However, the BMD increase was significantly lower with increasing age in the spine. No age dependency was observed in the hips. In general the effect of treatment was improved with increasing dosage and duration of treatment from 6 to 36 months.