Many genetically determined craniosynostosis syndromes feature limb anomalies, implying that pathways of cranial suture and limb morphogenesis share some identical components. Identification of heterozygous mutations in FGFR1, FGFR2, FGFR3, TWIST and MSX2 in craniosynostosis has focused particular attention on these genes. Here we explore two themes: use of clinical/molecular analysis to provide new clues to pathophysiology and the contrasting effects of loss- and gain-of-function mutations. Apert syndrome is a severe craniosynostosis/syndactyly disorder usually caused by specific substitutions (Ser252Trp or Pro253Arg) in FGFR2. The relative severity of cranial and limb malformations varies in opposite directions for the two mutations, suggesting that these phenotypes arise by different mechanisms. Clinical and biochemical evidence supports a model in which alternative splice forms of FGFR2 mediate these distinct effects. Pro-->Arg substitutions equivalent the Pro253Arg/FGFR2 mutation occur in both FGFR1 and FGFR3, and are also associated with craniosynostosis. This suggests a common pathological mechanism, whereby enhanced affinity for a limited repertoire of tissue-specific ligand(s) excessively prolongs signalling in the cranial suture. The first MSX2 mutation in craniosynostosis was described in 1993 but this remains the only example. We have recently identified three MSX2 mutations associated with a different cranial phenotype, parietal foramina. DNA binding studies show that the craniosynostosis and parietal foramina arise from gain and loss of function, respectively.