The standard model of osteoclastogenesis uses M-CSF and RANKL to differentiate osteoclasts from peripheral blood mononuclear cells (PBMCs). Using real time PCR, we found that RANKL profoundly up-regulates the GM-CSF receptor in this model. This suggests that GM-CSF receptor might represent a target for regulation, whereby osteoclast precursors integrate RANKL and GM-CSF signals to either promote or inhibit differentiation. Exogenous GM-CSF was added to the in vitro osteoclastogenesis model to test these alternative hypotheses. Cells were examined through time, with TRAP staining, morphology and gene array analysis. Continuous exposure to GM-CSF totally represses osteoclast differentiation; and resulted in an alternative cell phenotype induced by GM-CSF in the presence of M-CSF and RANKL when compared to the alternative treatments. A 19,000 gene microarray was used to compare GM-CSF+RANKL+M-CSF treated cells against osteoclasts. Microarray analysis showed GM-CSF mediated repression of osteoclast differentiation was concurrent with suppression of osteoclast marker genes: cathepsin K; osteoclast specific H+ ATPase; surface marker, CD68; and transcription factors we have shown are up-regulated in osteoclasts, such as NFATc1. The inhibition by GM-CSF of known osteoclast markers indicates an alternative GM-CSF dependent differentiation pathway. Real-time PCR analysis of 7 regulated genes validated the array data (FBP, GABPa, GABPb, ILF3, Kox31, NFATc1 and SCYA2). These data support the hypothesis that GM-CSF receptor up-regulation by RANKL sensitises the cell for inhibition of differentiation. In the cytokine milieu of the bone marrow, the ratio of GM-CSF and RANKL may be an important determinant of osteoclast differentiation.