We propose a numerical means of increasing the objectivity of describing, characterizing, and evaluating craniofacial morphology, including dysmorphology. A craniofacial variability index (CVI) can be created for an individual by obtaining a series of anthropometric measurements of the head and face, converting each measurement value to a standardized z-score, and then calculating the standard deviation for the whole set of these z-score values. This value is lowest for those faces in which all of the z-scores are in the same direction and of relatively the same magnitude and highest for faces with pronounced difference in direction and size of the individual z-scores. The CVI has a range of values with an approximately normal distribution in a reference population of 1,312 individuals. Examination of a small sample of individuals with known craniofacial syndromes indicates that the CVI in such cases consistently falls outside of the normal range of the index, and its value is highest in individuals with the greatest subjective degree of facial involvement. Finally, the CVI is robust: age, sex, size of the individual, and changes in the number of variables used in its calculation have little impact on its distribution. When used in conjunction with traditional clinical assessment, the CVI has a number of potential clinical applications including initial diagnostic screening, demonstrating age-related changes in postnatal development of patients with facial syndromes, and pre-and post-surgical assessments of indi-viduals with craniofacial anomalies.