Despite the relatively low frequency of cervical spine fractures in trauma patients, tremendous resources are expended on the use of imaging to exclude fracture. Some level 2 evidence can direct the selection of subjects for imaging and optimization of the imaging strategy. A suggested algorithm for evidence-based cervical spine imaging is shown in Fig. 1. This algorithm is based on the sequential assessment of two questions: (1) Is imaging necessary? (2) If imaging is necessary, what is the optimal strategy? The NEXUS and the Canadian cervical spine prediction rule investigations are large methodologically sound observational studies of clinical indications for cervical spine imaging that have addressed the question of who should undergo imaging. The results of these studies indicate that simple clinical criteria can be used to exclude fracture safely without imaging in many low-risk subjects. Data from these studies suggest that the implementation of such prediction rules into practice may reduce unnecessary imaging, although more research is necessary to document the actual effects. In subjects in whom imaging is indicated, cost-effectiveness analysis can be performed to determine the optimal imaging strategy. For high-risk subjects, cost-effectiveness analysis suggests that CT is the preferred initial strategy. When compared with radiography, the higher short-term costs of CT are counter-balanced by the decreased need for further imaging in patients without injury and by the increased sensitivity for fracture. The high-risk cervical spine criteria used at the author's center seem to be valid for identifying appropriate patients for initial imaging with CT.