Anticonvulsant therapy causes changes in calcium and bone metab olism and may lead to decreased bone mass with the risk of osteo porotic fractures. The two widely used antiepileptic drugs pheny toin and carbamazepine are recognized to have direct effects on bone cells. The aim of our study was to measure the influence of long-term treatment with antiepileptic drugs on bone mineral den sity (BMD) and to look on direct effects of carbamazepine and phenytoin on human osteoblast-like cells.BMD was measured by dual-energy X-ray absorptiometry. Mark ers of bone formation and bone resorption were determined in serum and urine. Data of 59 patients were compared to 55 age and sex matched controls. Direct effects of phenytoin and carbama zepine on human osteoblast-like cells were investigated in exper imental studies.BMD in the lumbar spine region (L2 through L4) was significantly lower in the patient group as compared to controls (p < 0.0004). At femoral sites BMD was lower in patients, but this difference did not reach statistical significance. The decrease in BMD at both sites was dependent on the duration of therapy. Excretion of pyridinoline crosslinks was markedly increased in the patients. 25-hydroxyvitamin D3 and 1,25-dihydroxy-vitamin D3 were significantly de creased in patients. Proliferation rate of human osteoblast-like cells was increased by phenytoin in low doses. Both, phenytoin and car bamazepine inhibited cell growth at concentrations equivalent to therapeutic doses for the treatment of epileptic diseases. Our clinical and experimental data indicate that long-term treat ment with anticonvulsant drugs leads to a lower BMD. The exper imentally observed decrease in bone cell proliferation might be clinically associated with impaired new bone formation. Beside alterations in calcium and vitamin D homeostasis leading to osteomalazia, direct effects of anticonvulsant drugs on bone cells may contribute to the damaging effects on the skeletal system.