Some osteopetrotic mutations lead to low resorption, increased numbers of osteoclasts, and increased bone formation, whereas other osteopetrotic mutations lead to low resorption, low numbers of osteoclasts, and decreased bone formation. Elaborating on these findings, we discuss the possibility that osteoclasts are the source of anabolic signals for osteoblasts. In normal healthy individuals, bone formation is coupled to bone resorption in a tight equilibrium. When this delicate balance is disturbed, the net result is pathological situations, such as osteopetrosis or osteoporosis. Human osteopetrosis, caused by mutations in proteins involved in the acidification of the resorption lacuna (ClC‐7 or the a3‐V‐ATPase), is characterized by decreased resorption in face of normal or even increased bone formation. Mouse mutations leading to ablation of osteoclasts (e.g., loss of macrophage‐colony stimulating factor [M‐CSF] or c‐fos) lead to secondary negative effects on bone formation, in contrast to mutations where bone resorption is abrogated with sustained osteoclast numbers, such as the c‐src mice. These data indicate a central role for osteoclasts, and not necessarily their resorptive activity, in the control of bone formation. In this review, we consider the balance between bone resorption and bone formation, reviewing novel data that have shown that this principle is more complex than originally thought. We highlight the distinct possibility that osteoclast function can be divided into two more or less separate functions, namely bone resorption and stimulation of bone formation. Finally, we describe the likely possibility that bone resorption can be attenuated pharmacologically without the undesirable reduction in bone formation.