Objective The current classification of mandibular condyle fractures as basal, low neck, and high neck as reported by Loukota et al. (Br J Oral Maxillofac Surg 43:72-73, 2005) and Neff et al. (Craniomaxillofac Trauma Reconstr 7:S44-S58, 2014) has a weakness. Nearly no high-neck fractures are reported (they are typically classified as type C head fractures) contrary to basal condylar fractures, which are overestimated (nearly all low-neck fractures are classified as basal). The aim of this study is to present a modified AO/SORG classification of mandibular condyle fractures. Material and methods A new arrangement of the reference lines is proposed because the fracture lines are mainly oblique in this region. The proposed classification was validated using a series of 84 cases that were treated surgically. Results The diagnoses using the proposed new classification system significantly differed from those based on the old system (p < 0.005). All basal fractures in the new classification system were also classified as basal in the old system. The same was true for type C head fractures. The differences were found for low-neck fractures (4 of 84 diagnoses differed between the old and new classifications, i.e., they were previously classified as basal fractures) and high-neck fractures (3 of 84 fractures were diagnosed as low-neck fractures or type C head fractures using the old classification). Conclusion The epidemiology of the condyle injury should be based on a classification, which reveals types of fractures which are represented by factually and frequently observed cases. That is why a relatively common AO/SORG classification can be modified for the benefit of assessing incidences of high-neck and low-neck fractures. Clinical relevance Considering that the treatment of the high-neck fractures is much technically complicated than the low-neck ones, this will have an influence on the management of trauma to the area.