Objective To evaluate the efficacy of internal fixation with or without fusion in the treatment of thoracolumbar burst fractures.Methods Clinical controlled trails related to the application of pedicle screw instrumentation with or without fusion for thoracolumbar fractures before March,2012were obtained by searching PubMed,Science Direct,Medline and CNKI.Quality evaluation was made on the included literature,from which data were extracted to integrate various rescarch results by using RevMan 5.1.The quantitative data were analyzed based on the effect scale of mean difference (MD) and bilateral 95％ confidence interval (CI).The numeration data were analyzed in the use of effect scale of odds ratio (OR) and bilateral 95％ CI.The merging of some data was manually completed.Results After retrieving,eight English and one Chinese papers of the clinical controlled trials,and two related Meta analysis were obtained.With exclusion of one repetitive research,eight papers were involved in the review.Meta analysis demonstrated that fusion and non-fusion fixation had no significant differences in aspects of correction of kyphotic angle,correction and correction loss of vertebral body height,neurological function improvement,complication rate,and length of hospital stay.While compared with the fusion fixation,non-fusion fixation showed a more serious correction loss of kyphotic angle,a fewer blood loss and a shorter operation time.Conclusions Non-fusion fixation shows the similar efficacy with fusion fixation in the treatment of some thoracolumbar burst fractures pertaining to releasing compression,restoring spinal stability and preventing complications,but it can also significantly decrease operation time and blood loss.Furthermore,non-fusion fixation may markedly improve patients' quality of life since it restores motion of the instrumented segment after removal of implant and decreases the risk of adjacent segmental degeneration.