AbstractSex estimates is a key step of biological profile assessment in a forensic or anthropologic context. In this study, the sexual dimorphism of the frontal bone was analyzed to assess the accuracy of sex estimates using a geometric morphometric approach in a pre-pubertal and post-pubertal sample. The shape of the frontal bone was digitized on the lateral cephalograms of 87 pre-pubertal subjects (42 males, mean age 10.14, SD ± 1.48 years; 45 females mean age 10.02, SD ± 1.11 years) and 103 post-pubertal ones (53 males, mean age 29.33 SD ± 11.88 years; 50 females, mean age 26.77 SD ± 11.07 years). A generalized Procrustes analysis (GPA) was performed for shape analyses, filtering the effects of position, rotation, translation, and size. A principal component analysis (PCA) was performed on the GPA transformed variables, and a multiple logistic regression model was used to assess the accuracy of sex estimates. In both age groups, the average size of the centroid was significantly larger in males. The females presented shapes with a shorter distance between P2 (glabella) and P1 (supratoral) and a general narrowing of the structure on the sagittal plane. In the pre-pubertal group, the shape difference was not statistically significant. In the post-pubertal group, the mean shape was significantly different between the sexes. The method displayed a high accuracy for sex estimates (88.7% males, 90.3% females) also when applied in a validation sample (82.6% males and 94.1% females). The described morphometric analysis of the frontal bone is based on a limited number of landmarks, which allows sex estimates with high accuracy in post-pubertal subjects, while it is not applicable in pre-pubertal ones.