To test the hypothesis that during treatment with denosumab osteomorphs and precursors recycle to higher number of osteoclasts with time, we measured TRAcP5b in serum taken 6 months after the last injection in postmenopausal women treated for 1-10 years. Serum TRAcP5b values were not related to time of exposure to denosumab. Purpose In women with postmenopausal osteoporosis the aetiology of the observed inverse relationship between duration of denosumab (Dmab) therapy and bone loss after its discontinuation is currently unknown. In studies in mice inhibition of RANKL is associated with an increase in osteomorphs and osteoclast precursors that recycle into osteoclasts and may accumulate with time. We hypothesized that longer inhibition of RANKL by Dmab will be followed by the synchronous formation of a larger number of osteoclasts after stopping treatment. To test this hypothesis, we measured serum TRAcP5b, a marker of osteoclast numbers, in postmenopausal women treated with Dmab for different periods of time up to 10 years. Methods TRAcP5b, C-terminal telopeptide of type 1 collagen (CTX) and procollagen type 1 N-terminal propeptide (P1NP) were measured at 6.0 months ± 15 days after last Dmab injection in 59 women who had received Dmab for 4.0 ± 2.3 years (range 1-10 years). Of these, 38 were treatment naïve (group 1) and 21 had received other treatments prior Dmab (group 2). Results Duration of Dmab treatment was not related to serum TRAcP5b values or to TRAcP5b/CTX ratio either in the whole cohort or in each of the two groups separately. In contrast, serum TRAcP5b values were significantly correlated with serum CTX values (r s = 0.619; p < 0.001), but not with serum P1NP values or BMD at all skeletal sites. Conclusion Our observations indicate that serum TRAcP5b, measured at 6 months after a Dmab injection, is not a useful early marker for time-dependent increased accumulation of osteoclasts in humans and for identification of patients at risk for a higher rebound increase in bone resorption.