OBJECTIVE To assess the outcomes associated with C1-C2 transarticular screw fixation. METHODS The clinical outcomes of 121 patients treated with posterior C1-C2 transarticular screws and wired posterior C1-C2 autologous bone struts were evaluated prospectively. Atlantoaxial instability was caused by rheumatoid arthritis in 48 patients, C1 or C2 fractures in 45, transverse ligament disruption in 11, os odontoideum in 9, tumors in 6, and infection in 2. RESULTS Altogether, 226 screws were placed under lateral fluoroscopic guidance. Bilateral C1-C2 screws were placed in 105 patients; each of 16 patients had only one screw placed because of an anomalous vertebral artery (n = 13) or other pathological abnormality. Postoperatively, each patient underwent radiography and computed tomography to assess the position of the screw and healing. Most screws (221 screws, 98%) were positioned satisfactorily. Five screws were malpositioned (2%), but none were associated with clinical sequelae. Four malpositioned screws were reoperated on (one was repositioned, and three were removed). No patients had neurological complications, strokes, or transient ischemic attacks. Long-term follow-up (mean, 22 mo) of 114 patients demonstrated a 98% fusion rate. Two nonunions (2%) required occipitocervical fixation. In comparison, our C1-C2 fixations with wires and autograft (n = 74) had an 86% union rate. CONCLUSIONS Rigidly fixating C1-C2 instability with transarticular screws was associated with a significantly higher fusion rate than that achieved using wired grafts alone. The risk of screw malpositioning and catastrophic vascular or neural injury is small and can be minimized by assessing the position of the foramen transversaria on preoperative computed tomographic scans and by using intraoperative fluoroscopy and frameless stereotaxy to guide the screw trajectory.