Three-dimensional soft tissue facial morphometry was investigated in a sample of 167 children aged 6 to 9 years by using a new noninvasive computerized method. For each child, 16 cutaneous facial landmarks were automatically collected by a system consisting of two infrared CCD cameras, real-time hardware for the recognition of markers, and software for the three-dimensional reconstruction of the x, y, and z coordinates of landmarks. From these landmarks, 15 linear and 10 angular measurements and five linear distance ratios were computed. For each age class, mean values were computed for all children with a bilateral Angle Class I occlusion (modified according to Katz) and compared with values obtained in children with a bilateral Class II occlusion. Most of the differences involved three-dimensional angular measurements: Class II children had more convex faces in the sagittal plane and a less prominent mandible than did Class I children. No differences were found in the linear measurements. Only the lower facial height ratio was different between the two occlusion groups, but the difference was not consistent among all the age groups.