Several studies suggest that 24,25-dihydroxyvitamin D [24,25(OH)₂D] may have an effect on bone mass and metabolism. We evaluated the relationship between serum 24,25(OH)₂D levels and bone density and bone metabolism in children with a primary bone disorder-osteogenesis imperfecta (OI). The study included 132 patients (age, 1.1 to 17.9 yr; 67 girls) with OI types I, III, or IV who had not received bisphosphonate treatment at the time of analysis. Serum 24,25(OH)₂D levels were significantly higher in OI type III than in OI type I or IV. Serum 24,25(OH)₂D concentrations were positively correlated with serum 25-hydroxyvitamin D (25OHD) levels and negatively correlated with serum PTH levels, and were not correlated with serum 1α,25-dihydroxyvitamin D [1,25(OH)₂D]. The ratio between serum 24,25(OH)₂D and 25OHD was negatively correlated with age and was independent of serum 25OHD concentrations. Regression analysis revealed that OI severity (P = 0.04), serum 25OHD levels (P < 0.001), and serum PTH concentrations (P = 0.045), but not age, gender, or serum 1,25(OH)₂D, were independent predictors of serum 24,25(OH)₂D levels. No correlation was found between serum 24,25(OH)₂D levels or the ratio between serum 24,25(OH)₂D and 25OHD and lumbar spine bone mineral density z-scores or bone marker levels (serum osteocalcin and urinary collagen type I N-telopeptide) after adjusting for OI type, age, and gender. Patients with more severe OI type had higher 24,25(OH)₂D serum levels and higher serum 24,25(OH)₂D to 25OHD ratios, suggesting an increased 25OHD-24-hydroxylase activity.