Premature closure of the sagittal suture occurs as an isolated (nonsyndromic) birth defect or as a syndromic anomaly in combination with other congenital dysmorphologies. The genetic causes of sagittal nonsyndromic craniosynostosis (NSC) remain unknown. Although variation of the dysmorphic (scaphocephaly) skull shape of sagittal NSC cases has been acknowledged, this variation has not been quantitatively studied three-dimensionally (3D). We have analyzed the computed tomography skull images of 43 infants (aged 0.9-9 months) with sagittal NSC using anatomical landmarks and semilandmarks to quantify and characterize the within-sample phenotypic variation. Suture closure patterns were defined by dividing the sagittal suture into three sections (anterior, central, posterior) and coding each section as 'closed' or 'fused'. Principal components analysis of the Procrustes shape coordinates representing the skull shape of 43 cases of NSC did not separate individuals by sex, chronological age, or dental stages of the deciduous maxillary first molar. However, analysis of suture closure pattern allowed separation of these data. The central section of the sagittal suture appears to be the first to fuse. Then, at least two different developmental paths towards complete fusion of the sagittal suture exist; either the anterior section or the posterior section is the second to fuse. Results indicate that according to the sequence of sagittal suture closure patterns, different craniofacial complex shapes are observed. The relationship between craniofacial shape and suture closure indicates not only which suture fused prematurely (in our case the sagittal suture), but also the pattern of the suture closure. Whether these patterns indicate differences in etiology cannot be determined with our data and requires analysis of longitudinal data, most appropriately of animal models where prenatal conditions can be monitored.