The aim of this study was to quantify the relative contributions of genetic and environmental factors to variations in dental arch breadth, length and palatal height in a sample of Australian twins, and to estimate heritabilities using modern model-fitting methods. Dental casts of 20 male and 24 female monozygous (MZ) twin pairs, 17 male and 8 female dizygous (DZ) twin pairs, and 9 opposite-sexed DZ twin pairs were selected from the collection of records of twins housed at the Adelaide Dental School. The mean ages of subjects were 15.8 +/- 3.5 years (MZ) and 17.0 +/- 4.7 years (DZ). Dental casts were scanned using a contact-type 3D scanner, PICZA interfaced to a personal computer running 3D-Rugle3 software. Data were subjected to univariate genetic analysis with the structural equation modelling package, Mx, using the normal assumptions of the twin model. A model incorporating additive genetic (A) and unique environmental (E) variation was found to be the most parsimonious for dental arch breadth and length, and palatal height. Estimates of heritability for dental arch breadth ranged from 0.49 to 0.92, those for arch length from 0.86 to 0.94, and those for palatal height were 0.80 and 0.81, respectively. These results indicate a high genetic contribution to the variation in dental arch dimensions in mainly teenage twins.