We are satisfied from studies with mi mi osteopetrotic mutant mice that osteoclasts arise from the myeloid tissue of bone marrow and not as formerly proposed from osteoprogenitor cells. Grafts of compatible normal myeloid cells cure the osteopetrosis by the substitution of the qualitatively defective osteoclasts with normal ones. Nevertheless it is still not fully clear through what cellular cascade this is effected. Current opinion would favour the pathway from pluripotent haematopoietic stem cells to circulating monocytes to tissue macrophages with ultimate fusion to form multinucleate osteoclasts. However, it is recorded that osteoclasts differ from macrophage polykaryons of inflammatory tissue not only in certain subcellular characteristics but in absence of Fc and C3 receptors. We can explain this as due to development through a specialised line of osteoclast precursors independent of conventional macrophages, if current unpublished experimental studies confirm the transfer to osteoclasts of the additional "beige" marker incorporated into grafted material.