Although most children survive B cell acute lymphoblastic leukemia (B-ALL), they frequently experience long-term, treatment-related health problems, including osteopenia and osteonecrosis. Because some children present with fractures at ALL diagnosis, we considered the possibility that leukemic B cells contribute directly to bone pathology. To identify potential mechanisms of B-ALL–driven bone destruction, we examined the p53−/−; Rag2−/−; Prkdcscid/scid triple mutant (TM) mice and p53−/−; Prkdcscid/scid double mutant (DM) mouse models of spontaneous B-ALL. In contrast to DM animals, leukemic TM mice displayed brittle bones, and the TM leukemic cells overexpressed Rankl, encoding receptor activator of nuclear factor κB ligand. RANKL is a key regulator of osteoclast differentiation and bone loss. Transfer of TM leukemic cells into immunodeficient recipient mice caused trabecular bone loss. To determine whether human B-ALL can exert similar effects, we evaluated primary human B-ALL blasts isolated at diagnosis for RANKL expression and their impact on bone pathology after their transplantation into NOD.Prkdcscid/scidIl2rgtm1Wjl/SzJ (NSG) recipient mice. Primary B-ALL cells conferred bone destruction evident in increased multinucleated osteoclasts, trabecular bone loss, destruction of the metaphyseal growth plate, and reduction in adipocyte mass in these patient-derived xenografts (PDXs). Treating PDX mice with the RANKL antagonist recombinant osteoprotegerin–Fc (rOPG-Fc) protected the bone from B-ALL–induced destruction even under conditions of heavy tumor burden. Our data demonstrate a critical role of the RANK-RANKL axis in causing B-ALL–mediated bone pathology and provide preclinical support for RANKL-targeted therapy trials to reduce acute and long-term bone destruction in these patients.