Skeleton bones, distinguished by trabecular and cortical bone tissue content, exhibit varied growth and composition, in response to modified dietary calcium and phosphorus levels. The study investigated how gilts adapt their individual bone and bone region mineralisation kinetics in response to changing intake of Ca and P. A total of 24 gilts were fed according to a two-phase (Depletion (D) 60–95 and Repletion (R) 95–140 kg BW, respectively). During the D phase, gilts were fed either 60% (D60) or 100% (D100) of the estimated P requirement. Subsequently, during the R phase, half of the gilts from each D diet were fed either 100% (R100) or 160% (R160) of the estimated P requirement according to a 2 2 factorial arrangement. Bone mineral content (BMC) was assessed in the whole body, individual bones (femur and lumbar spine L2–L4), and bone regions (head, front legs, trunk, pelvis, femur, and hind legs) every 2 weeks using dual-energy X-ray absorptiometry (DXA). At 95 kg BW, gilts fed D60 showed reduced BMC and BMC/BW ratio in all studied sites compared to those fed D100 (P < 0.001). During the depletion phase, the allometric BW-dependent regressions slopes for BMC of D100 gilts remained close to 1 for all sites and did not differ from each other. In contrast, the slopes were lower in D60 gilts (P < 0.05), with an 18% reduction in the whole body, except for the front and hind legs, femur, and pelvis, which exhibited higher reductions (P < 0.05). At 140 kg BW, BMC and BMC/BW ratio of all studied sites were similar in gilts previously fed D60 and D100, but higher in R160 than in R100 gilts (P < 0.05), except for front and hind legs. During the repletion phase, the allometric BW dependent regressions slopes for BMC were lower (P < 0.05) in R100 than in R160 gilts (for whole body -10%; P < 0.01) except for front and hind legs, femur, and pelvis. In conclusion, bone demineralisation and recovery followed similar trends for all measured body sites. However, the lumbar spine region was most sensitive whereas the hind legs were least sensitive. These data suggest that using bone regions such as the head and forelegs that can be collected easily at the slaughterhouse may be a viable alternative to whole body DXA measurement.