Purpose We aimed to comprehensively analyse a possible correlation between skeletal malocclusions, gender and mandibular characteristics in all three dimensions in adults and to identify mandibular characteristics that are typical for extreme skeletal patterns. Methods A 3D model of the skull was calculated in 111 adult patients (mean age = 27.0 +/- 10.2 years; 49 women, 62 men) from available computed tomography or cone beam computed tomography scans of their heads. Based on the 3D models, the skeletal patterns were examined in (a) the transversal dimension regarding asymmetry according to menton deviation, (b) the sagittal dimension according to the Wits appraisal and (c) the vertical dimension according to the maxillomandibular plane angle. The mandibular characteristics assessed were linear (ramus height and width, body length), angular (ramus, gonial and body angle) and volumetric (ramus/mandibular volume, body/mandibular volume) parameters. Results No correlation between transversal skeletal asymmetry and mandibular characteristics were found, while sagittal (F(16, 174) = 3.32, p < 0.001, eta(2) = 0.23) and vertical (F(16, 174) = 3.18, p < 0.001, eta(2) = 0.23) skeletal patterns were shown to have a significant effect on the mandible. Gender correlated with mandibular characteristics independently from the skeletal pattern. Discriminant analysis revealed that class II and III patients differed in ramus and body angle with class II patients showing higher angles (ramus angle: class II = 89.8 +/- 3.9 degrees vs. class III = 84.4 +/- 4.8 degrees; body angle: class II = 87.7 +/- 4.8 degrees vs. class III = 82.1 +/- 5.2 degrees). Hypo- and hyperdivergent patients were discriminated by gonial angle, body angle and body/mandibular volume with hyperdivergent patients having a greater gonial and body angle and body/mandibular volume (gonial angle: hypodivergent = 114 +/- 9.3 degrees vs. hyperdivergent = 126.4 +/- 8.6 degrees; body angle: hypodivergent = 82.9 +/- 4.4 degrees vs. hyperdivergent = 87.7 +/- 6.5 degrees; body/mandibular volume: hypodivergent = 72.4 +/- 2.7% vs. hyperdivergent = 76.2 +/- 2.6%). Conclusion When analysing 3D data for treatment planning of adult patients, the orthodontist should pay attention to angular and volumetric characteristics of the mandible to identify extreme skeletal sagittal or vertical malocclusions.