The purpose of this paper is to provide a new perspective on craniosynostosis by correlating what is known about sutural biology with the events of craniosynostosis per se. A number of key points emerge from this analysis: (1) Sutural initiation may take place by over lapping, which results in beveled sutures, or by end-to-end approximation, which produces nonbeveled, end-to-end sutures. All end-to-end sutures occur in the midline (e.g., sagittal and metopic) probably because embryonic biomechanical forces on either side of the initiating suture tend to be equal in magnitude. A correlate appears to be that only synostosed sutures of the midline have pronounced bony ridging. (2) Long-term histologic observations of the sutural life cycle call into question the number of layers within sutures. The structure varies not only in different sutures, but also within the same suture over time. (3) Few, if any, of the many elegant experimental research studies in the field of sutural biology have increased our under standing of craniosynostosis per se. An understanding of the pathogenesis of craniosyn ostosis requires a genetic animal model with primary craniosynostosis and molecular techniques to understand the gene defect. This may allow insight into pathogenetic mechanisms involved in primary craniosynostosis. It may prove to be quite heterogeneous at the basic level. (4) The relationship between suture closure, cessation of growth, and functional demands across sutures poses questions about various biological relationships. Two conclusions are provocative. First, cessation of growth does not necessarily, or always lead to fusion of sutures. Second, although patent sutures aid in the growth process, some growth can take place after suture closure. (5) In an affected suture, craniosynostosis usually begins at a single point and then spreads along the suture. This has been shown by serial sectioning and calls into question results of studies in which the affected sutures are only histologically sampled. (6) Craniosynostosis is etiologically and pathogenetically heterogeneous. Known human causes are reviewed. Is craniosynostosis simply normal suture closure commencing too early? In hyperthyroidism, this is probably the case, but in Apert syndrome, true sutures in the sagittal and metopic regions fail to form ab initio. The actual mechanisms that result in pathologic synostosis, although incompletely understood, must be heterogeneous in nature. (7) Special topics are analyzed and discussed including fetal head constraint, the calvaria in Apert syndrome and holoprosencephaly, a critique of Moss's theory, calcified cephalohematoma, secondary cartilage, paradoxical craniosynostosis, and delayed suture closure. (8) The effect of craniosynostosis on the calvaria and the cranial base is discussed; it is shown that the relationship between the calvaria and the cranial base varies and reflects several different factors such as whether or not the basal portion of the coronal ring is involved and whether the synostosis is isolated or syndromic. (9) Simple craniosynostosis involving the coronal suture may produce significant effects on the midface. Restricted coronal suture growth is the primary effect, changes in the cranial base the secondary effect, and foreshortening of the midface the tertiary effect. The effects follow a temporal sequence. In general, the degree of midface shortening is a function of which suture is restricted from growing, how early growth restriction of the suture takes place, and how much time elapses before measuring the effects on the midface. This model of craniofacial change is only applicable to simple craniosynostosis, not to cases with complete coronal ring involvement or to syndromic cases such as Apert or Crouzon syndromes. © 1993 Wiley-Liss, Inc.