Background Bone cement leakage is a major complication in the treatment of percutaneous vertebroplasty for Kümmell’s disease, and the focus of close attention during the surgery. The purpose of this article was to investigate the clinical outcomes of Kümmell’s disease treated by percutaneous vertebroplasty with or without bone cement leakage. Methods A total of 64 patients with Kümmell’s disease from December 2016 to February 2018 treated by percutaneous vertebroplasty were included in the study. After the treatment, 32 cases were respectively divided into two groups according to X-ray examination of bone cement leakage: leakage group and non-leakage group. Preoperative course, age, sex, bone mineral density, damaged segment, anterior vertebral height, vertebral compression rate, Cobb angle, visual analogue scale and Oswestry dysfunction index were compared between the two groups. After surgery, the amount of bone cement injected, operation time, adjacent vertebral refracture rate, visual analogue scale, Oswestry dysfunction index, the recovery value of vertebral anterior height and the improvement value of Cobb angle were compared between the two groups. Results The course, age and Cobb angle of the leakage group were significantly greater than those of the non-leakage group (P< 0.05, respectively). The height of anterior vertebral margin and bone mineral density in the leakage group were significantly lower than those in the non-leakage group (P< 0.05, respectively). The two groups were followed up for at least 24 months. The amount of bone cement injected was significantly greater in the leakage group than in the non-leakage group (P=0.000). Visual analogue scale and Oswestry dysfunction index of the two groups on the second day after surgery and at the last follow-up were significantly lower than these before surgery (P< 0.05, respectively), but there was no significant difference between the two groups. In the leakage group, the recovery value of the anterior edge height of the injured vertebra and the improvement value of the Cobb angle on the second day after surgery and at the last follow-up were significantly improved compared with the non-leakage group (P< 0.05, respectively). Conclusion Percutaneous vertebroplasty is an effective and minimally invasive treatment for Kümmell’s disease. The leakage group had longer course, older age, more serious kyphotic deformity, vertebral compression and osteoporosis, and higher amount of bone cement injected than these of the non-leakage group. However, there were not significant differences in the rate of adjacent vertebral refractures, visual analogue scale and Oswestry dysfunction index between the two groups. Therefore, the bone cement leakage does not affect the surgical effect.