OBJECTIVE To determine the efficacy of flexion-extension (F/E) cervical-spine radiographs in detecting acute cervical-spine instability in emergency patients. METHODS We retrospectively reviewed the charts of 141 consecutive trauma patients who had F/E views performed after a routine cervical-spine series (three views) was obtained in the emergency department. Interpretations of the routine series were compared with those of the F/E views to determine if additional useful information was provided by the latter. The charts also were reviewed to determine if any variables were associated with an increased use of F/E views, an increased likelihood of these views demonstrating instability, or any neurologic sequelae resulted from these studies. METHODS An urban Level I adult trauma center. METHODS The interpretations of the routine series were noted to be either normal, abnormal but without demonstrable fracture/dislocation, or demonstrating a fracture/dislocation. The F/E views were categorized as stable, unstable, or uninterpretable. RESULTS Cervical-spine instability was demonstrated by F/E views in 11 of the 141 patients (8%), four of whom had normal routine cervical-spine films. Three of these four patients required surgical stabilization. Prolonged neck pain (more than 24 hours), an initially abnormal spine series, and a neurosurgical consult were all associated with an increased use of F/E views. Ten of 11 patients with radiographic instability had significant neck pain by history; the remaining patient was intoxicated. No neurologic sequelae resulted from performing F/E studies. There was one false-negative F/E study, which raises concern about the reliability of this procedure in the ED. CONCLUSIONS We believe that a large prospective study is required to determine which patients warrant F/E views.