Orthodontists usually assume that genetic factors are of major significance to the development of malocclusion. Studies of twin similarities have lent support to this concept. Use of the twin model to estimate genetic variance, however, involves the assumptions that there are no mean or total variance inequalities between monozygotic and dizygotic twins. The environmental (including biologic) differences between monozygotic and dizygotic twin development may therefore constitute a source of bias. Genetic variance and heritability were estimated for a series of arch and occlusal traits in sixty twins using recently developed methods that are unbiased by variance heterogeneity between zygosities. Heterogeneity was demonstrated for twenty of the forty traits, suggesting considerable amounts of hidden environmental determinance. Tooth-displacement means differed between monozygotes and dizygotes, indicating bias in that trait's genetic variance analysis. Significant heritability could not be demonstrated for overbite, overjet, buccal segment relation, total tooth displacement, and occlusal discrepancies in arch shape. Arch size, individual tooth displacement scores, and cross-bite show significant genetic variance, averaging only about 36 percent of the total variance of these traits.