Technical advances are radically altering our concepts of normal prenatal craniofacial development. These include concepts of germ layer formation, the establishment of the initial head plan in the neural plate, and the manner in which head segmentation is controlled by regulatory (homeobox) gene activity in neuromeres and their derived neural crest cells. There is also a much better appreciation of ways in which new cell associations are established. For example, the associations are achieved by neural crest cells primarily through ceil migration and subsequent cell interactions that regulate induction, growth, programmed cell death, etc. These interactions are mediated primarily by two groups of regulatory molecules, "growth factors" (e.g., FGF and TGFoc) and the so-called steroid/thyroid/retinoic acid superfamily. Considerable advances have been made with respect to our understanding of the mechanisms involved in primary and secondary palate formation, such as growth, morphogenetic movements, and the fusion/merging phenomenon. Much progress has been made on the mechanisms involved in the final differentiation of skeletal tissues.Molecular genetics and animal models for human malformations are providing many insights into abnormal development. A mouse model for the fetal alcohol syndrome (FAS), a mild form of holoprosencephaly, demonstrates a mid-line anterior neural plate deficiency which leads to olfactory placodes being positioned too close to the mid-line, and other secondary changes. Work on animal models for the retinoic acid syndrome (RAS) shows that there is major involvement of neural crest cells. There is also major crest cell involvement in similar syndromes, apparently including hemifacial microsomia. Later administration of retinoic acid prematurely and excessively kills ganglionic placodal cells and leads to a malformation complex virtually identical to the Treacher Collins syndrome. Most clefts of the lip and/or palate appear to have a multifactorial etiology. Genetic variations in TGFocs, RARocs, NADH dehydrogenase, an enzyme involved in oxidative metabolism, and cytochrome P-450, a detoxifying enzyme, have been implicated as contributing genetic factors. Cigarette smoking, with the attendant hypoxia, is a probable contributing environmental factor. It seems likely that few clefts involve single major genes. In most cases, the pathogenesis appears to involve inadequate contact and/or fusion of the facial prominences or palatal shelves. Specific mutations in genes for different FGF receptor molecules have been identified for achondroplasia and Crouzon's syndrome, and in a regulatory gene (Msx2) for one type of craniosynostosis. Poorly co-ordinated control of form and size of structures, or groups of structures (e.g., teeth and jaws), by regulatory genes should do much to explain the very frequent "mismatches" found in malocclusions and other dentofacial "deformities".Future directions for research, including possibilities for prevention, are discussed.