Three-dimensional computed tomography (CT) has an important role in determining the presence and extent of congenital and acquired craniofacial deformities. The authors compared the sensitivity and specificity of three-dimensional CT in the detection and characterization of craniosynostosis with that of planar CT and skull radiography. Eighty-two patients with isolated and syndromal synostoses were imaged with CT and three-dimensional CT, and 42 with skull radiography. Three-dimensional CT scan processing was performed by shaded-surface reconstruction, volumetric, and depth-coded methods. Two trained observers read each scan series in a blinded fashion. Diagnostic utility of the images was determined with receiver operating characteristic analysis. The observers ranked three-dimensional shaded images higher than the other types, with three-dimensional volumetric images second and three-dimensional surface images ranked third. Results of this study demonstrate that three-dimensional shaded-surface reconstruction from CT scans is superior to conventional plain radiographs and CT scans in diagnosing craniosynostosis.