Object. The management of odontoid fractures remains controversial. Only direct anterior screw fixation provides immediate stabilization of the spine and may preserve normal C1-2 motion. To determine the indications, optimum timing, and results for direct anterior screw fixation of odontoid fractures, the authors reviewed the surgery-related outcome of patients who underwent this procedure at two institutions. Methods. One hundred forty-seven consecutive patients (98 males and 49 females) who underwent direct anterior screw fixation for recant (less than or equal to 6 months postinjury [129 patients]) or remote (greater than or equal to 18 months postinjury [18 patients]) Type II (138 cases) or III (nine cases) odontoid fractures at the University of Utah (94 patients) and National Institute of Traumatology in Budapest, Hungary (53 patients) between 1986 and 1998 are included in this study (mean follow up 18.2 months). Data obtained from clinical examination, review of hospital charts, operative findings, and imaging studies were used to analyze the surgery-related results in these patients. In patients with recent fractures there was an overall bone fusion rate of 88%. The rate of anatomical bone fusion of recent fractures was significantly (p less than or equal to 0.05) higher in fractures oriented in the horizontal and posterior oblique direction (compared with anterior oblique), but this finding was independent (p greater than or equal to 0.05) of age, sex, number of screws placed (one or two), and the degree or the direction of odontoid displacement. In patients with remote fractures there was a significantly lower rate of bone fusion (25%). Overall, complications related to hardware failure occurred in 14 patients (10%) and those unrelated to hardware in three patients (2%). There was one death (1%) related to surgery. Conclusions. Direct anterior screw fixation is an effective and safe method for treating recent odontoid fractures (less than or equal to 6 months postinjury). It confers immediate stability, preserves C1-2 rotatory motion, and achieves a fusion rate that compares favorably with alternative treatment methods. In contradistinction, in patients with remote fractures (greater than or equal to 18 months postinjury) a significantly lower rate of fusion is found when using this technique, and these patients are believed to be poor candidates for this procedure.