Background: Coflex, a kind of dynamic interspinous spacer, has been widely used for the treatment of lumbar spinal stenosis in the past a few years. However, controversy remains as to whether dynamic interspinous spacer use is superior to traditional decompression and fusion surgery. High complication, reoperation rate, and costs with poor outcomes were also mentioned in the literature. The aim of the present study is to evaluate whether Coflex implantation following spinal decompression provided better clinical outcomes compared with traditional decompression and fusion for symptomatic lumbar spinal stenosis through midterm follow-up. Methods and findings: A total of 100 patients who were confirmed L4/5 lumbar spinal stenosis was surveyed from June 2007 to June 2010. They were randomly and equally divided into two groups: 50 cases underwent spinal decompression with Coflex implantation, and 50 cases were treated with spinal decompression with fixation and fusion. The operation time, intraoperative blood loss, ambulation time, and hospitalization days, Japanese Orthopedic Association scores, visual analogue scale scores, Oswestry disability index and SF-36 scores were compared between the two groups. The ranges of motion and the height loss at adjacent segments (L3/4 and L5/S1) were measured preoperative and postoperative, respectively. Adjacent segment degeneration at L3/4 and L5/Sl was assessed by Pfirrmann classification. Complications were also recorded. The average age was 57.6 ± 5.9 years old in Coflex implantation group and 59.0 ± 6.7 years old in fusion group, respectively. The average follow-up period was 7.12 ± 1.1 year in Coflex implantation group and 7.31 ± 1.6 year in fusion group, respectively. JOA, ODI, VAS and SF-36 scores were improved at the last follow-up in all the two groups with significant differences (P 0.05). The intervertebral heights of adjacent segments were decreased at the last follow-up and the ranges of intervertebral motions were increased in both groups. The height loss and the range of motion increase of adjacent segments were greater in fusion group than those in Coflex group with statistical significant difference (P<0.01). At the last follow-up, adjacent segment disc Pfirrmann grade progressed more obviously in fusion group compared with that in Coflex group, and there was significant difference (P<0.05) between the two groups. Conclusion: Based on the present study, it showed that Coflex implantation and fusion after spinal decompression had the same clinical outcomes and satisfaction in treatment of symptomatic lumbar spinal stenosis after 7 years follow-up. Nevertheless, Coflex implantation had the advantages of less bleeding loss, less trauma and quick recovery. Compared with fusion surgery, Coflex implantation had also advantages in maintaining intervertebral height and delaying intervertebral disc degeneration of adjacent segments.