We studied osteoclastic differentiation from normal and osteopetrotic human CD14 cells in vitro. Defects in acid transport, organic matrix removal, and cell fusion with deficient attachment were found. Analysis of genotypes showed that TCIRG1 anomalies correlated with acid transport defects, but surprisingly, organic matrix removal failure correlated with CLCN7 defects; an attachment defect had normal TCIRG1 and CLCN7. Introduction: Osteopetrotic subjects usually have normal macrophage activity, and despite identification of genetic defects associated with osteopetrosis, the specific developmental and biochemical defects in most cases are unclear. Indeed, patients with identical genotypes often have different clinical courses. We classified defects in osteoclast differentiation in vitro using four osteopetrotic subjects without immune or platelet defects, three of them severe infantile cases, compared with normals. Materials and Methods: Osteoclast differentiation used isolated CD14 cells; results were correlated with independent analysis of two key genes, CLCN7 and TCIRG1. CD14 cell attachment and cell surface markers and extent of differentiation in RANKL and colony-stimulating factor (CSF)-1 were studied using acid secretion, bone pitting, enzyme, and attachment proteins assays. Results and Conclusions: CD14 cells from all subjects had similar lysosomal and nonspecific esterase activity. With the exception of cells from one osteopetrotic subject, CD14 cells from osteopetrotic and control monocytes attached similarly to bone or tissue culture substrate. Cells from one osteopetrotic subject, with normal CLCN7 and TCIRG1, did not attach to bone, did not multinucleate, and formed no podosomes or actin rings in RANKL and CSF-1. Attachment defects are described in osteopetrosis, most commonly mild osteopetrosis with Glantzman's thrombasthenia. However, this case, with abnormal integrin αvβ3 aggregates and no osteoclasts, seems to be unique. Two subjects were compound heterozygotes for TCIRG1 defects; both had CD14 cells that attached to bone but did not acidify attachments; cell fusion and attachment occurred, however, in RANKL and CSF-1. This is consistent with TCIRG1, essential for H+-ATPase assembly at the ruffled border. A compound heterozygote for CLCN7 defects had CD14 cells that fused in vitro, attached to bone, and secreted acid, TRACP, and cathepsin K. However, lacunae were shallow and retained demineralized matrix. This suggests that CLCN7 may not limit H+-ATPase activity as hypothesized, but may be involved in control of organic matrix degradation or removal.