Leposternon microcephalum is a species belonging to the Amphisbaenia, a group of burrowing reptiles. Amphisbaenia present various morphological and physiological adaptations that allow them to penetrate the ground and live underground, through a system of galleries and permanent chambers that they build themselves. Among the morphological adaptations in this group, those of the skull stand out as it serves as the main excavation tool. Four basic skull shapes are recognized: rounded, keeled, shovel‐shaped, and spade‐shaped. The skull of L. microcephalum belongs to this last type, which is considered the most specialized. The species inhabits soils that are highly compacted and difficult to penetrate. Among the species of Leposternon present in South America, L. microcephalum has the widest distribution, being found in all Brazilian biomes and neighboring countries such as Bolivia, Argentina, Paraguay, and Uruguay. The analysis of the skull of this species was carried out using three‐dimensional geometric morphometrics (3D‐GMM), a technique that allows comparative analysis, through robust statistical methods, of shape and its variations, using Cartesian coordinate data from a configuration of homologous landmarks. The technique allows the size and shape components of a structure to be analyzed separately. From an ontogenetic point of view, this methodology had also been used to investigate variations in Cynisca leucura, a member of the Amphisbaenidae with a rounded head. Our hypothesis is that the patterns of morphological differentiation in the skull, mainly in the intermediate and occipital regions, are similar in different Amphisbaenia species. Therefore, the objective of this study was to analyze cranial morphological variations in an ontogenetic series of L. microcephalum using 3D‐GMM. Computed Tomographic scans of 13 specimens were analyzed: juveniles (N = 8) and adults (N = 5), based on 20 landmarks that characterize the skull. Principal components and regression analyses between shape (dependent variable) and size (independent variable) showed a clear difference between the cranial morphological pattern of juvenile individuals and that of adults. For instance, young specimens tend to have a dorsoventrally tall neurocranium, with the tip of the snout more anteriorly oriented and its dorsal border subtly curved. Dorsally, the parietal region is thicker and smoothly dome‐shaped in juveniles. As in C. leucura, the variation was strongly correlated with the size change from juvenile to adult, indicating a dominant role for ontogenetic allometry in determining skull shape.