We analyzed retrospectively 27 individuals with C-2, C-3 fractures, 8 of whom were treated with a Philadelphia collar rather than the more commonly used halo vest. Successful fusion without neurological deficit occurred in all cases whether treatment was by Philadelphia collar (n = 8), halo apparatus (n = 16), or prolonged bed rest (n = 3). There was an increase in subluxation in 3 patients in both the Philadelphia collar group and the halo apparatus group. However, 1 of the 3 in the collar group was uncooperative and discarded his collar during treatment. If he is excluded, the subluxation rate in the Philadelphia collar group is 27%; the rate in the halo group is 19%. The Philadelphia collar seems to be an acceptable means of stabilizing the neck in C-2, C-3 fractures, but one must consider the degree of cervical movement in the collar (as detected by dynamic x-ray films), patient reliability, and patient age. The presence of subluxation did not preclude successful fusion in either the Philadelphia collar or the halo vest groups.