Metabolic bone disease is common in children and adults with chronic cholestasis. We evaluated baseline vitamin D (vitamin Dz and D3), 25-OH vitamin Dz and D3, 1,25(OH), vitamin D, vitamin D-binding protein, bone mineral content and dietary mineral content in six children (mean age: 12.1 years) with cholestasis since infancy. Absorption of 25-OH vitamin DS and vitamin Dz was evaluated by measuring serial serum concentrations after a test dose. Bone mineral content was reduced by >2 S.D. in five of six subjects compared to agespecific controls; none had radiographic evidence of rickets but all had osteopenia. Dietary Ca and P content in the subjects was comparable to the recommended daily allowance for age-specific children. Baseline serum vitamin Dz concentrations were undetectable in all but one cholestatic subject despite oral supplementation with 2,500 to 50,000 IU per day vitamin Dz. Baseline serum 25-OH vitamin D was 33.2 f 6.0 ng per ml (mean f S.E.) and comparable to our laboratory norms (15 to 50 ng per ml). Serum 1,25(0H)z vitamin D and "free" 1,25(0H)z vitamin D were both significantly (p < 0.05) reduced compared to controls. A significantly blunted rise and reduced area under the absorption curve (both p < 0.001) after 1,000 IU per kg vitamin D, was found in cholestatic children (0.8 ng f 0.5 ng per ml and 18.0 2 14.3 ng h r per ml, respectively) compared to controls (59.5 f 10.0 ng per ml and 1,780 f 253 ng hr per ml, respectively). The peak change and area under the absorption curve for 25-OH vitamin DS from baseline after 10 pg per kg 25-OH vitamin DB was significantly reduced (both p < 0.05) in cholestatic children (60.2 f 13.8 ng per ml and 751.5 f 189.9 ng hr per ml, respectively) compared to controls (151.6 f 30.7 ng per ml and 2,021 2 472 ng h r per ml, respectively). We conclude that despite severe vitamin D malabsorption, normal serum 25-OH vitamin D concentrations in cholestatic children most likely result from sunlightstimulated endogenous vitamin D synthesis. Use of oral