Reports of abnormalities in vitamin D, calcium, and bone status associated with anticonvulsant use are inconsistent and difficult to interpret because of widely varying study designs, particularly for ambulatory status. We examined the relative effects of anticonvulsant use and ambulatory status on vitamin D, calcium, and bone status in a large group (n = 338) of children who had either normal motor function (ambulatory) or were nonambulatory and either receiving anticonvulsants or not; all had developmental delays. Data included diet records, serum analyses (calcium and calcidiol), and hand-wrist radiographs evaluated for bone maturation and quality. Data were analyzed by using a general linear models (GLM) procedure. Dietary and biochemical data were compared with those of a group of 34 normal children. There were no differences in calcium or vitamin D intakes among the four study groups; however, a high percentage of intakes was below the recommended dietary allowances for calcium (56%) and vitamin D (70%). Vitamin D intakes were positively associated with serum calcium (P < 0.005) and calcidiol (P < 0.01) concentrations. Analysis of covariance indicated that ambulatory status but neither anticonvulsant use nor their interaction contributed significantly to the prediction of serum calcium (P < 0.009) and calcidiol (P < 0.0001), the Z scores for number of ossified centers (P < 0.008), bone age (P < 0.0001), and bone area (P < 0.003). A strong interaction between anticonvulsant use and ambulatory status was seen for percentage cortical area (P < 0.0008), which was entirely due to anticonvulsant use in nonambulatory children (effect size = 0.98). Results suggest that ambulatory status is more important than was recognized previously in relation to abnormalities in vitamin D, calcium, and bone statuses; that all nonambulatory children may be at risk for low serum calcidiol and osteopenia; and that routine monitoring of risk and consideration of prophylactic vitamin D supplementation are warranted.