Introduction Kyphoplasty for osteoporotic vertebral compression fractures was short but painful. The purpose of the current study was to investigate the effects and safety of ultrasound-guided thoracic paravertebral block in patients undergoing percutaneous kyphoplasty (PKP). Methods A prospective study of 195 patients with thoracic compression fracture undergoing PKP was conducted. The patients were non-randomly assigned to receive an ultrasound-guided thoracic paravertebral block (Group P, n=96) and local infiltration anesthesia (Group L, n=99), and were compared along with intraoperative parameters and anesthetic effects. Visual analog scale (VAS) of pain, systolic blood pressure (SBP), diastolic blood pressure (DBP), and heart rate (HR) were measured at pre-anesthesia (T0), post-anesthesia (T1), trocar insertion (T2), balloon dilatation (T3), cement injection (T4) and post-operation (T5). Results The anesthesia time was significantly longer in group P than in group L. There were no significant differences in the epidemiological data, operation time, blood loss, hospitalization time complications and costs between the two groups. The additional analgesics rate was significantly lower in group P than in group L. The investigators’ satisfaction scores, patients’ anesthesia satisfaction scores and intention rate of re-administration anesthesia were significantly higher in group P than in group L. Intraoperative VAS score (T1-T4) was significantly lower in group P than in group L. There was no significant difference in VAS scores (T0 and T5), SBP, DBP and HR between these two groups. Conclusion Ultrasound - guided thoracic paravertebral block achieved more effective analgesia in comparison to local infiltration groups in patients undergoing percutaneous kyphoplasty.