Thirty-five relevant studies involving in total 682 patients with 709 different types of injuries were evaluated in a review to determine the outcomes after immobilisation in a halo vest for various injuries to the upper cervical spine between 1962 and 1998. Studies were analysed according to the type of injury pattern and in terms of the treatment outcomes following primary treatment with a halo vest. The following types of injuries were evaluated: odontoid fractures (n = 420), hangman's fractures (n = 172), other axis fractures (n = 75), Jefferson fractures (n = 26), C1 arch fractures (n = 9), atlantooccipital (n = 2) and atlantoaxial dislocations (n = 5). The ligamentary atlantooccipital dislocations never healed. All isolated Cl ring fractures healed completely. The isolated C1 arch fractures healed in 83% of the cases. The ligamentary atlantoaxial dislocations had a 60% rate of healing. Healing was noted in all isolated odontoid type I fractures, 85% of the isolated odontoid type II fractures, and 67% of the odontoid type II fractures with combined injuries. The isolated odontoid type III fractures had a 97% healing rate. The non-classifiable odontoid fractures had a healing rate of 85%. The stable C2 arch fractures (hangman's fracture) healed consistently in 99%, and 90% success was found for other C2 fractures. A halo vest can be recommended for patients with isolated Jefferson fractures, hangman's fractures, odontoid type III and type II fractures, with a low dislocation rate. The results of treatment with a halo vest were unsatisfactory with regard to combined injuries with an odontoid type II fracture. An overall healing rate of 86%, however, allows one to conclude that this treatment continues to be a good alternative to operative stabilisation of bone injuries to the upper cervical spine.