Purpose: To compare panoramic and cone beam computed tomography (CBCT) determinations of implant-to-mandibular canal (MC) dimensions in mandibular regions posterior to the mental foramen and to investigate whether factors such as gender, age, region, and vertical dimension influence correlation between the two techniques. Materials and Methods: A retrospective analysis was carried out in 64 consecutive adult patients (42 females, 22 males; average age 57.1 ± 13.3 years) in whom 126 implants were positioned in the posterior segment of the mandible. Implant sites (first premolar, second premolar, first molar, and second molar) were assessed on each panoramic and CBCT radiograph by measuring the distance from the inferior border of the implant to the superior border of the MC. Binary logistic regression analysis was used to compute the odds ratios (ORs) of each implant site for underestimation vs nonunderestimation. Linear regression analysis was performed with CBCT dimension as the dependent variable and panoramic dimension, gender, and age as the independent variables. Results: The mean implant-to-MC dimension was 2.50 ± 1.31 mm in panoramic radiography and 2.91 ± 1.62 mm in CBCT. The OR that an implant at the second molar region belonged to the underestimation group was strong (15.1:1) and highly significant (P = .011). If a predictive value of .95 was demanded, the implant-to-MC dimensions had to be overestimated by 2 mm compared to the predicted CBCT dimension. Conclusion: This study provides evidence of an underestimation of available vertical bone dimensions for implants in the posterior regions of the mandible when assessed by panoramic radiography. Use of CBCT is therefore recommended for all implant size estimations in this region.