For medico-legal forensic practitioners the identification of unknown remains is an important part of any investigation, often predicated on having accurate estimations of age and sex. In considering the specific skeletal elements available to facilitate such biological information, the cranium is frequently targeted for analysis, as it exhibits marked traits of sexual dimorphism, and also has a predictable pattern of growth. There are, however, instances where it may not be possible to estimate skeletal sex, especially in the juvenile skeleton. There is still considerable uncertainty surrounding the age at which the human cranium is quantifiably dimorphic. The aim of the present study is to explore age and sex variation in three-dimensionally reconstructed MDCT scans of the juvenile cranium. The study sample comprises 152 juvenile crania from a Western Australian population; a total of 52 three-dimensional landmarks are acquired and analyzed using Procrustean geometric morphometrics. Group discrimination is assessed between sexes and across age classes. Results demonstrate that sexual dimorphism and age variation is discernible through geometric morphometric analysis of form, size and shape. Relative to sex and age, size is found to be generally equivalent to, or even more accurate than, shape data. There is little quantifiable sexual dimorphism in individuals younger than 12 years of age with most variation related to age; discrimination improves with increasing age, with average hit rate (HR) values increasing from just over 50% (52–58%) to more than 90% (93–94%) accuracy at 18 years. In contrast, differences between contiguous age classes follow the opposite trend and tend to be larger in prepubertal groups, while becoming progressively smaller in older age classes. This study demonstrates that simple linear interlandmark distances describing overall cranial size may provide a simple option for preliminary classifications of age and sex in skeletal remains of forensic interest. However, although recombining size and shape to perform analyses using form generally does not appreciably improve predictive accuracy, it potentially contributes to increased confidence in group assessment (especially for sex) and thus offers a promising, albeit complex, type of information to discriminate groups based on cranial size and/or shape.