Background Orbital decompression is an established procedure for the management of exophthalmos and visual rehabilitation from optic neuropathy in cases of thyroid eye disease. Numerous procedures for removal of orbital bony wall, fat or a combination of these for a variety of indications in different stages of the disease have been well reported in the medical literature. However, the relative effectiveness and safety of these procedures in relation to the various indications remains unclear. Objectives To review current published evidence for the effectiveness of surgical orbital decompression for disfiguring proptosis in adult thyroid eye disease and summa rise information on possible complications and the quality of life from the studies identified. Search methods We searched CENTRAL (which contains the Cochrane Eyes and Vision Group Trials Register) (The Cochrane Library 2011, Issue 10), MEDLINE (January 1950 to October 2011), EMBASE (January 1980 to October 2011), the metaRegister of Controlled Trials (mRCT) (www.controlled-trials.com) and ClinicalTrials.gov (http://clinicaltrials.gov). There were no date or language restrictions in the electronic searches for trials. The electronic databases were last searched on 6 October 2011. We searched oculoplastic textbooks, conference proceedings from the European and American Society of Ophthalmic Plastic and Reconstructive Surgery (ESOPRS, ASOPRS), European Ophthalmological Society (SOE), the Association for Research in Vision and Ophthalmology (ARVO) and American Academy of Ophthalmology (AAO) for the years 2000 to 2009 to identify relevant data. We attempted to contact researchers who are active in this field for information about further published or unpublished studies. Selection criteria We included randomised controlled trials (RCTs) with no restriction on date or language comparing two or more surgical methods for orbital decompression with removal of bony wall, orbital fat or a combination of both for disfiguring proptosis or comparison of surgical techniques with any form of medical decompression. Data collection and analysis Each review author independently assessed study abstracts identified from the electronic and manual searches. Author analysis was then compared and full papers for appropriate studies were obtained according to the inclusion criteria. Disagreements between the authors were resolved by discussion. Main results We identified two randomised trials eligible for inclusion in the review. There was significant variability between the trials for interventions, methodology and outcome measures and therefore meta-analysis was not performed. One study suggested that the transantral approach and endoscopic transnasal technique had similar effects in reducing exophthalmos but that the endoscopic approach may be safer, relating to fewer complications. This study had short-term follow-up and lacked information on our primary outcome (success or failure of treatment). The second study provided evidence that intravenous steroids may be superior to primary surgical decompression in the management of compressive optic neuropathy requiring less secondary surgical procedures, although it relates more frequently to transient side effects. This study was weakened by a small sample size. Until more credible evidence is available recommendations as to best treatment cannot be reliably made. Authors' conclusions A single study showed that the transantral approach for orbital decompression was related to more complications than the endoscopic transnasal technique which is preferred by Ear, Nose and Throat (ENT) surgeons, usually as an adjunctive procedure. Intravenous steroids were reported in a single trial to be the most efficient intervention for dysthyroid optic neuropathy. The majority of published literature on orbital decompression for thyroid eye disease consists of retrospective, cohort, or case series studies. Although these provide useful descriptive information, clarification is required to show the relative effectiveness of each intervention for various indications. The two RCTs reviewed are not robust enough to provide credible evidence to our understanding of current decompressive surgery and to support recommendations for clinical practice. There is evidence from currently available uncontrolled studies that removal of the medial and lateral wall (balanced decompression) with or without fat removal may be the most effective surgical method related to only a few complications. There is a clear need for randomised studies evaluating the balanced two-wall, three-wall and orbital fat decompression techniques. Comparison with other surgical techniques for orbital decompression or with immunosuppression in cases of compressive optic neuropathy would also be important. These studies should primarily address the reduction of exophthalmos, disease severity, complication rates, quality of life and cost of the intervention.