Background Anterior instability is a frequent complication following a traumatic glenohumeral dislocation. Frequently the underlying pathology associated with recurrent instability is a Bankart lesion. Surgical correction of Bankart lesions and other associated pathology is the key to successful treatment. Open surgical glenohumeral stabilisation has been advocated as the gold standard because of consistently low postoperative recurrent instability rates. However, arthroscopic glenohumeral stabilisation could challenge open surgical repair as the gold standard treatment for traumatic anterior glenohumeral instability. Objectives Primary evidence that compared the effectiveness of arthroscopic versus open surgical glenohumeral stabilisation was systematically collated regarding best-practice management for adults with traumatic anterior glenohumeral instability. Search strategy A systematic search was performed using 14 databases: MEDLINE, Cumulative Index of Nursing and Allied Health (CINAHL), Allied and Complementary Medicine Database (AMED), ISI Web of Science, Expanded Academic ASAP, Proquest Medical Library, Evidence Based Medicine Reviews, Physiotherapy Evidence Database, TRIP Database, PubMed, ISI Current Contents Connect, Proquest Digital Dissertations, Open Archives Initiative Search Engine, Australian Digital Thesis Program. Studies published between January 1984 and December 2004 were included in this review. No language restrictions were applied. Selection criteria Eligible studies were those that compared the effectiveness of arthroscopic versus open surgical stabilisation for the management of traumatic anterior glenohumeral instability, which had more than 2 years of follow up and used recurrent instability and a functional shoulder questionnaire as primary outcomes. Studies that used non-anatomical open repair techniques, patient groups that were specifically 40 years or older, or had multidirectional instability or other concomitant shoulder pathology were excluded. Data collection and analysis Two independent reviewers assessed the eligibility of each study for inclusion into the review, the study design used and its methodological quality. Where any disagreement occurred, consensus was reached by discussion with an independent researcher. Studies were assessed for homogeneity by considering populations, interventions and outcomes. Where heterogeneity was present, synthesis was undertaken in a narrative format; otherwise a meta-analysis was conducted. Results Eleven studies were included in the review. Two were randomised controlled trials. Evidence comparing arthroscopic and open surgical glenohumeral stabilisation was of poor to fair methodological quality. Hence, the results of primary studies should be interpreted with caution. Observed clinical heterogeneity in populations and outcomes was highlighted and should be considered when interpreting the meta-analysis. Authors also used variable definitions of recurrent instability and a variety of outcome measures, which made it difficult to synthesise results. When comparable data were pooled, there were no significant differences (P > 0.05) between the arthroscopic and open groups with respect to recurrent instability rates, Rowe score, glenohumeral external rotation range and complication rates. Conclusions Statistically, it appears that both surgical techniques are equally effective in managing traumatic anterior glenohumeral instability. In light of the methodological quality of the included studies, it is not possible to validate arthoscopic stabilisation to match open surgical stabilisation as the gold standard treatment. Further research using multicentred randomised controlled trials with sufficient power and instability-specific questionnaires with sound psychometric properties is recommended to build on current evidence. The choice of treatment should be based on multiple factors between the clinician and the patient.