Objective: Given the increasing incidence of deformational plagiocephaly due to infants’ supine sleeping position to prevent sudden infant death syndrome, reliable anthropometric diagnostics are needed. Besides the traditional method of measuring landmarks with callipers, three-dimensional (3D) photography has great potential. In this investigation the accuracy of 3D photogrammetry is studied. Methods: The study included 100 randomly chosen children between the ages of 4 and 20 months with a non-synostotic cranial deformity in a retrospective analysis. Measurements of diagonals A and B on the infant's head were obtained once using callipers. 3D photographs of these children were measured 5 times by 5 clinicians separately. Results: The inter- and intra-rater agreements of the 3D measurements had low variability in the variance component analysis. The standard deviations for reproducibility and repeatability were 0.117–0.283 cm for diagonals A and B. The intra-class correlation coefficients for the inter-rater reliability resulted in excellent agreement (0.97 for plagiocephaly, 0.98 for brachycephaly, 0.96 for combined deformity). The comparison of the 3D photographic and callipers measurements showed that 3D photography resulted in a slight over-estimation. Conclusion: 3D photogrammetry is potentially a reliable tool for treatment planning and follow-up of abnormal head shapes in infancy.