Prostate cancer metastases to bone are observed in around 80% of prostate cancer patients and represent the most critical complication of advanced prostate cancer, frequently resulting in significant morbidity and mortality. As the underlying mechanisms are not fully characterized, understanding the biological mechanisms that govern prostate cancer metastases to bone at the molecular level should lead to the determination of new potential therapeutic targets. Receptor activator of NFkappaB ligand (RANKL)/RANK/Osteoprotegerin (OPG) are the key regulators of bone metabolism both in normal and pathological condition, including prostate cancer bone metastases. In the present study, we demonstrated that human prostate cancer cell lines, DU145 and PC3 express biologically functional RANK. Indeed, soluble human RANKL (shRANKL, 100 ng/ml) treatment induced ERK 1/2, p38 and IkappaB phosphorylations in these cells. shRANKL administration also promoted DU145 and PC3 prostate cancer cell invasion in vitro. Whereas human OPG (hOPG) administration alone (100 ng/ml) had no marked effect, combined association of both agents abolished the RANKL-induced DU145 cell invasion. As RANKL had no direct effect on DU145 cell proliferation, the observed effects were indeed related to RANKL-induced cell migration. DU145 human prostate cancer cells promoted osteoclastogenesis of osteoclast precursors generated from mouse bone marrow. Moreover, DU145 cells produced soluble factor(s) that up-regulate the proliferation of MC3T3-E1 pre-osteoblasts through the activation of the ERK 1/2 and STAT3 signal transduction pathways. This stimulation of pre-osteoblast proliferation resulted in an increased local RANKL expression that can activate both osteoclasts/osteoclast precursors and prostate cancer cells, thus facilitating prostate cancer metastasis development in bone. We confirm that RANKL is a factor that facilitates metastasis to bone by acting as an activator of both osteoclasts and RANK-positive prostate cancer cells in our model. Furthermore, the present study provides the evidence that blocking RANKL-RANK interaction offer new therapeutic approach not only at the level of bone resorbing cells, but also by interfering with RANK-positive prostate cancer cells in the prostate cancer bone metastasis development.