Background Arthroscopic decompression is commonly used to treat shoulder impingement, while the role of conservative treatment in these patients remains unclear. The objective of this paper is to synthesise available evidence regarding the effectiveness of subacromial decompression for shoulder impingement compared with conservative treatment to verify whether arthroscopic surgery is superior to conservative treatment. Methods We systematicly retrieved the Cochrane databases, Embase as well as Pubmed (from inception to July. 02, 2019) for randomized controlled trials. Cochrane risk-of-bias tool was used to assess all referred studies’ quality and we pooled outcomes with a random-effects model. We divided the outcomes into short-term subgroup(<2 years) and long-term subgroup (≥2 years). Results 7 randomised controlled trials (RCTs) were included, involving a total of 607 patients, 297 patients operated arthroscopic decompression, compared to 310 patients treated with conservative management. We found no significant differences either in shoulder pain scores or shoulder function scores between arthroscopic decompression and conservative management wether in long-term or short-term follow-up subgroups. Conclusion No significant difference was demonstrated on the treatment outcomes of shoulder impingement between arthroscopic decompression and conservative management in our meta-analysis. Thus, we suggested that conservative management should be chosen firstly for patients with shoulder impingement, when patients’ symptom cannot be relieved from the conservative treatment, arthroscopic decompression should be taken into consideration.Levels of Evidence Level-I study