This study investigated the predictive value of orthopantomography (OPG) for the difficulty of extracting impacted mandibular third molars, in comparison with cone-beam computed tomography (CBCT). In this descriptive quantitative investigation, two oral and maxillofacial radiologists evaluated OPG and three-dimensional CBCT images according to the Pell-Gregory and Winter classifications. The results for the classification were compared using the chi-square test, and the prediction of difficulty was assessed using the Pederson scale, with a significance level of p< 0.05. The study included 30 patients (14 men and 16 women), providing a total of 53 samples of impacted mandibular third molars. Of these, 30 (56.6%) were from the right side and 23 (43.4%) from the left. There was a statistically significant difference between the OPG and CBCT images concerning their relation to the mandibular ramus (p< 0.05). However, evaluations based on occlusal lines and angulation showed no significant differences (p> 0.05). According to the Pederson scale, significant differences were observed between OPG and CBCT in predicting extraction difficulty (p< 0.05). CBCT offered a more accurate assessment of the surgical difficulty associated with mandibular third molars than OPG. OPG views frequently failed to adequately visualize the region of the mandibular ramus, influencing the perceived difficulty of mandibular third molar surgery. In certain cases, the use of CBCT imaging is crucial.