Increased bone turnover and rapid bone loss follow discontinuation of denosumab. We investigated the long-term efficacy of zoledronate (ZOL) in maintaining bone mineral density (BMD) after discontinuation of denosumab. In this randomized, open-label, interventional study, we included 61 postmenopausal women and men older than 50 years discontinuing denosumab after 4.6 ± 1.6 years. We administered ZOL 6 months (6 M) or 9 months (9 M) after the last denosumab or when bone turnover had increased (observation group [OBS]). ZOL was readministrated if p-cross-linked C-terminal telopeptide (p-CTX) increased ≥1.26 μg/L or BMD decreased ≥5%. The results after 12 months have previously been published;​ here we report the outcome after 24 months (ClinicalTrials NCT03087851). Fifty-eight patients completed the study. From 12 to 24 months after the initial ZOL, lumbar spine (LS) BMD was maintained:​ 0.9 ± 0.9%, 0.4 ± 0.8%, and 0.3 ± 0.7% in the 6 M, 9 M, and OBS groups, respectively (p > .05, no between-group differences). Similarly, total hip (TH) and femoral neck (FN) BMD did not change in any group during year 2. From baseline to 24 months after ZOL, LS BMD decreased by 4.0 ± 0.8%, 4.1 ± 0.8%, and 4.3 ± 1.5% in the 6 M, 9 M, and OBS groups, respectively (p < .001, no between-group differences). Significant bone loss (LS, TH, or FN) was found in all groups 24 months after ZOL:​ 6 M group:​ n = 12 (60%), 9 M group:​ n = 7 (37%), and OBS group:​ n = 10 (53%). P-CTX did not change significantly during the second year (p > .05, no between-group differences). No patient fulfilled the CTX or fracture criteria for retreatment during year 2;​ however, 9 patients were retreated at M24 due to BMD loss ≥5%. Two patients sustained a non-vertebral fracture during year 2. Treatment with ZOL subsequent to long-term denosumab did not fully prevent increased bone turnover and bone loss during the first year;​ however, CTX remained with the reference range and BMD was maintained during the second year.