Dental cone-beam computed tomography (CBCT) has been available for only about 15 years, but has already found many uses in the practice of oral surgery despite a limited evidence base for its diagnostic efficacy. As with all X-ray techniques, careful attention must be given to justification and optimisation. It is usually associated with higher radiation doses and financial costs than conventional radiographic techniques, yet has corresponding advantages over multislice computed tomography. Equipment varies considerably in image quality, and it is hard to generalise about its clinical use, although it usually has acceptable accuracy for linear measurements. In terms of clinical applications, much of the literature remains dominated by opinion and case reports, with little published work relating to patient outcome efficacy. In implant dentistry, its use is well established, and guideline documents exist which favour selected use. For mandibular third molar surgery, it has good diagnostic accuracy for inferior dental canal relationships but, apart from 'high-risk' cases, there is no evidence to support routine use. It has many other potential uses where threedimensional information will be of help. Use of CBCT has implications for training of users. At this time, the evidence base suggests that a restrained approach in using CBCT is appropriate, limiting it to situations where conventional radiography cannot provide the information required to manage a patient adequately, or to those where multislice computed tomography is the current imaging modality of choice and where limited soft tissue contrast is acceptable.