The conventional model that bisphosphonates bind to the bone surface and inhibit mature osteoclasts does not convincingly explain the prolonged duration of action of zoledronate. We hypothesized that zoledronate on the bone surface adjacent to marrow cells impairs osteoclastogenesis, contributing to sustained inhibition of resorption. In this case, numbers of circulating preosteoclasts may be reduced after zoledronate treatment. This study assessed this possibility in subjects from a clinical trial. Twenty-two osteopenic women participating in a randomized, controlled trial comparing zoledronate 5 mg with placebo were recruited, 18 months after administration of study drug. Peripheral blood mononuclear cells were analyzed for the presence of osteoclast precursors using flow cytometry for preosteoclast markers and the ability to form osteoclast-like cells in culture with RANKL and M-CSF. There was no difference in the percentage of CD14(+)/CD11b(+) cells in peripheral blood between the two groups. The numbers of TRAP(+) multinucleated cells in cultures in the absence of RANKL and M-CSF were very low in both groups, but a significantly higher number of these cells was observed in the zoledronate group compared with the placebo group (p = 0.01). The number of TRAP(+) multinucleated cells and resorption pits following culture with RANKL and M-CSF did not differ between the two groups. Serum P1NP was reduced 53 % at 18 months in the zoledronate group but unchanged in the placebo group. These results do not support the hypothesis that the inhibitory action of zoledronate contributes to its prolonged action on preosteoclasts within bone marrow.