A sandwich vertebra is formed after multiple osteoporotic vertebral fractures treated by percutaneous vertebroplasty, which has a risk of developing new fractures. The purpose of our study was to (i) investigate the occurrence of new fractures in sandwich vertebra after cement augmentation procedures and to (ii) evaluate the clinical outcomes after prophylactic vertebral reinforcement applied with resorbable bone cement. From June 2011 to 2014, we analysed 55 patients with at least one sandwich vertebrae and treated with percutaneous vertebroplasty. Eighteen patients were treated by prophylactic vertebroplasty with a resorbable bone cement to strengthen the sandwich vertebrae as the prevention group. The others were the non-prevention group. All patients were examined by spinal radiographs within 1 day, 6 months, 12 months, 24 months and thereafter. The incidence of sandwich vertebra is 8.25% (55/667) in our study. Most sandwich vertebrae (69.01%, 49/71) are distributed in the thoracic-lumbar junction. There are 24 sandwich vertebrae (18 patients) and 47 sandwich vertebrae (37 patients) in either prevention group or non-prevention group, respectively. No significant difference is found between age, sex, body mass index, bone mineral density, cement disk leakage, sandwich vertebrae distribution or Cobb angle in the two groups. In the follow-up, 8 out of 37 (21.6%) patients (with eight sandwich vertebrae) developed new fractures in non-prevention' group, whereas no new fractures were detected in the prevention group. Neither Cobb angle nor vertebral compression rate showed significant change in the prevention group during the follow-up. However, in the non-prevention group, we found that Cobb angle increased and vertebral height lost significantly (P<0.05). Prophylactic vertebroplasty procedure applied with resorbable bone cement could decrease the rate of new fractures of sandwich vertebrae.