Spinosaurid theropod dinosaurs appear to represent convergent morphological evolution toward a crocodylian-like cranial morphology, previously linked to the possibility that spinosaurs adopted a similar, partially piscivorous, trophic niche. Further conclusions are hindered by a lack of quantitative evidence, and an incomplete understanding of the functional significance of key crocodylian cranial characters. A comparative biomechanical analysis of function in the snout of the spinosaurid theropod Baryonyx walkeri has been performed, comparing B. walkeri with a generalised large theropod dinosaur and two extant crocodylians (Alligator, Gavialis) that represent different endpoints of extant crocodylian morphological diversity. The aims of the analysis were (a) to determine which group is the closest functional analogue to B. walkeri, and (b) investigate the mechanical influence on cranial function of the antorbital fenestra and the secondary palate; morphological characters that appear to be of importance in both crocodyliform and spinosaur functional morphology. Results demonstrate that the closest structural and biomechanical analogue to B. walkeri is the extant gharial, rather than the alligator or conventional theropods. The secondary palate confers strength to the alligator skull in torsion, but provides resistance to bending in gharials and B. walkeri. Loss of the antorbital fenestra strengthens narrow or tubular theropod and gharial snouts, but has limited influence on the broader-snouted alligator morphotypes. Consequently, with their large antorbital fenestrae and lack of secondary palate, most theropod skulls were surprisingly suboptimally constructed to resist feeding-related bite loads. The mechanical impetus for archosaur palatal development and fenestral closure appears more complex than previously thought.