Background: The present study aimed to develop an accurate method for evaluating hip-bone development in DDH. Methods: CT scans were performed on 15 phantoms and 16 unilateral hips with developmental dysplasia. The validity of the three-dimensional reconstruction (3DR) method and ellipsoid model (EM), as well as ossification volumes (OVs) and ossification-volume ratios (OVRs), were analyzed. Results: Mean +/- SD EM- and 3DR-absolute-errors were 20.02 +/- 23.67 mL and 0.25 +/- 2.72 mL. The SNR of the 3DR method was 0.129, 50 times greater than that of the EM. The OVs of hips in the dislocated group (8.62 +/- 4.09/cm(3), 3.93 +/- 1.94/cm(3), and 20.21 +/- 8.93/cm(3) for the ischium, pubis and proximal femur) were significantly smaller (P = 0.008, 0.002, 0.006) than those in the normal group (11.61 +/- 5.63/cm(3), 4.93 +/- 2.73/cm(3), and 23.42 +/- 12.48/cm(3), respectively) and for the ilium (32.84 +/- 17.36/cm(3) and 31.14 +/- 16.25/cm(3), respectively; P = 0.018). OVRs in the young group (age < 44 months; 1.03 +/- 0.33, 0.91 +/- 0.15, and 0.93 +/- 0.05 for the ischium, pubis and proximal femur, respectively) were significantly higher (P = 0.010, 0.018, and 0.023, respectively) than those in the old group (participants aged >= 44 months; 0.60 +/- 0.24, 0.75 +/- 0.09, and 0.84 +/- 0.08, respectively), except for the ilium (1.09 +/- 0.17 and 1.06 +/- 0.08, respectively, P = 0.644). Conclusion: The accuracy of the 3DR method is higher than that of EM. OVs of dislocated hips are smaller. OVRs are associated with age but not dislocation distance. Bone development is delayed in dislocated hips and tends to reduce with age.