Direct grafts of human cells into immunocompromised or cortisone-treated animals, either alone or within carrier materials, have been used with some success to assess the developmental capability of the grafted cells. However, identification of the donor or host origins of the generated tissue in such direct grafts is essential. In an alternative and extensively used experimental system, cells are cultured within the isolated environments of diffusion chambers, which are surgically implanted in appropriate hosts. This system allows the direct study of the cellular potentials for differentiation as host tissues are excluded. In the present study, human osteoprogenitor cell populations derived from trabecular bone explants or marrow suspensions of 3 patients (2 females aged 14 years and 1 male aged 27 years) were cultured in the absence or the continuous presence of dexamethasone (10 nmol/L). Cells were impregnated into porous hydroxyapatite ceramics before subcutaneous implantation, or placed within diffusion chambers before intraperitoneal implantation, in athymic mice. All subcutaneous implants of cells in ceramic showed morphological evidence for the formation of bone tissue. In the diffusion chambers it was found that both marrow- and bone-derived fibroblastic cells cultured in the absence of dexamethasone generally produced fibrous tissue only. When cultured in the continuous presence of dexamethasone (10 nmol/L), these cell populations produced similar osteogenic tissues with active osteoblasts, wide osteoid seams, and mineralized tissue, with cartilage toward the interior of the chamber. These results validate the diffusion chamber as an experimental system to study human osteogenesis using appropriately primed cell populations.