Radiographic imaging of the mandibular third molar (M3) is a valuable diagnostic tool. It provides information on tooth position, root morphology, and relations to surrounding anatomical structures that facilitates diagnosis and treatment planning. Three modalities are commonly used in dentistry: intraoral and panoramic radiography, and cone beam computed tomography (CBCT). Over time, panoramic radiography became a justified choice in most cases of M3 removal. In the last decade, a three-dimensional (3D) alternative, CBCT, has seen increasing use. The advantages of 3D come at the cost of higher radiation doses and societal expense. The four studies in this thesis originated from current knowledge gaps and clinical needs. The use of panoramic radiography for evaluating the M3 and its relation to the inferior alveolar nerve (IAN) is well-established practice; however, its application has declined over time, possibly due to the rising popularity of CBCT imaging. Thus, it is of interest to investigate the suitability of panoramic imaging of the M3. Choice of imaging modality is important from dose delivery and socioeconomic standpoints. A survey of clinician experiences and preferences in imaging modalities for pre-surgical analysis was needed. In diagnostic imaging, the common goal is to reduce radiation exposure to the lowest threshold that still delivers reliable diagnostic information. Due to the growing use of CBCT, there is a need to optimize exposure settings and use reliable measurement methods in dosimetric analysis. Nevertheless, for the M3, the present literature has neither evaluated a low-dose protocol in a clinical setting nor compared the standard method of measuring absorbed dose in dental CBCT with an alternative method. This thesis is based on the following papers: Paper I, an observational study, evaluated M3 characteristics and IAN relation, incidental findings, and image quality regarding patient positioning on 442 panoramic radiographs. Key findings: A majority of M3s were erupted and vertically positioned, regardless of age. The IAN was located inferior to the roots in just over half of the cases; an overlapping position was most common if the tooth was retained or semi-retained, or patient age less than 30 years. Frequent incidental findings were apical radiolucencies, idiopathic osteosclerosis, and tooth fragments. Patient positioning was suboptimal in one-third of the radiographs; common errors included patient placement posterior to the image layer and an upward-tilted head. Paper II was a web-based questionnaire sent to general dentists and to residents and specialists in oral and maxillofacial surgery. The survey comprised multiple-choice questions with four M3 cases depicted in images. Key findings: A majority of respondents received a report within 2 weeks of their CBCT referral and would read it and view the images before surgery; one-third did not. Panoramic radiographs were the preferred modality in pre-surgical planning; differences between professions were significant. Panoramic radiographs and CBCT were seen as facilitating treatment planning, and CBCT as also reducing post-operative complications. CBCT tended to be preferred in more complex M3 cases. Paper III was a clinical trial investigating an alternative, low-dose CBCT protocol with lower tube current than the default protocol. After justification for CBCT, 48 patients (62 M3s) referred for a pre-surgical investigation were recruited. Two scans of each site were made using the two protocols Key findings: No significant differences occurred in visibility of root and mandibular canal relationship and proximity, root morphology, and possible root resorption of the second molar. Visibility of the periodontal ligament was significantly better in default images. Subjective image quality in low-dose images (60% less radiation) was acceptable in most cases. Paper IV was a laboratory study comparing two dosimeters in a CBCT scan of the M3 region. Five slices of an anthropomorphic phantom were prepared with thermoluminescent dosimeters (TLD-100) at 75 sites and covered with Gafchromic film LD-V1. Key findings: Point dose measurements at all TLD sites correlated well with doses measured on film; agreement was better at lower doses. Minimum and maximum doses on film versus at TLD sites deviated greatly on all slices and for several organs. To conclude, panoramic radiography is useful in evaluating M3 and IAN relationships. Clinicians consider pre-surgical information to differ depending on imaging modality; modality preference seems to depend on case complexity. Optimized CBCT protocols can be used for the M3. Gafchromic film is a favorable alternative in dental CBCT dosimetry