V-ATPase-mediated acid secretion is required for osteoclast bone resorption. Osteoclasts are enriched in V-ATPase a3 and d2 subunit isoforms, and disruption of either of their genes impairs bone resorption. Using purified fusion proteins of a3 N-terminal domain (NTa3) and full-length d subunits we determined in a solid-phase binding assay that half-maximal binding of d1 or d2 to immobilized NTa3 occurs at 3.1 ± 0.4 or 3.6 ± 0.6 nM, respectively, suggesting equally high-affinity interactions. A high-throughput modification of this assay was then used to screen chemical libraries for a3–d2 interaction inhibitors, and luteolin, a naturally occurring flavonoid, was identified, with half-maximal inhibition at 2.4 ± 0.9 µM. Luteolin did not significantly affect NIH/3T3 or RAW 264.7 cell viability, nor did it affect cytokine-induced osteoclastogenesis of RAW 264.7 cells or bone marrow mononuclear cells at concentrations ≤40 µM. Luteolin inhibited osteoclast bone resorption with an EC50 of approximately 2.5 µM, without affecting osteoclast actin ring formation. Luteolin-treated osteoclasts produced deeper resorption pits, but with decreased surface area, resulting in overall decreased pit volume. Luteolin did not affect transcription, or protein levels, of V-ATPase subunits a3, d2, and E, or V1V0 assembly. Previous work has shown that luteolin can be effective in reducing bone resorption, and our studies suggest that this effect of luteolin may be through disruption of osteoclast V-ATPase a3–d2 interaction. We conclude that the V-ATPase a3–d2 interaction is a viable target for novel anti-resorptive therapeutics that potentially preserve osteoclast–osteoblast signaling important for bone remodeling. J. Cell. Biochem. 114: 929–941, 2013. © 2012 Wiley Periodicals, Inc.