More than 99 per cent of the roughly 58,000 living vertebrate species have jaws. This major clade, whose members are collectively known as gnathostomes ('jawed mouths'), made its earliest definitive appearance in the Silurian period, 444-416 million years (Myr) ago, with both the origin of the modern (crown-group) radiation and the presumptive invasion of land occurring by the end of the Devonian period (359 Myr ago). These events coincided with a major faunal shift that remains apparent today: the transition from Silurian ecosystems dominated by jawless fishes (agnathans) to younger assemblages composed almost exclusively of gnathostomes. This pattern has inspired several qualitative descriptions of the trophic radiation and ecological ascendance of the earliest jawed vertebrates. Here we present a quantitative analysis of functional variation in early gnathostome mandibular elements, placing constraints on our understanding of evolutionary patterns during this critical interval. We document an initial increase in functional disparity in the Silurian that stabilized by the first stage of the Devonian, before the occurrence of an Emsian (∼400 Myr ago) oxygenation event implicated in the trophic radiation of vertebrates. Subsequent taxonomic diversification during the Devonian did not result in increased functional variation; instead, new taxa revisited and elaborated on established mandibular designs. Devonian functional space is dominated by lobe-finned fishes and 'placoderms'; high disparity within the latter implies considerable trophic innovation among jaw-bearing stem gnathostomes. By contrast, the major groups of living vertebrates--ray-finned fishes and tetrapods--show surprisingly conservative mandibular morphologies with little indication of functional diversification or innovation. Devonian gnathostomes reached a point where they ceased to accrue further mandibular functional disparity before becoming taxonomic dominants relative to 'ostracoderm'-grade jawless fishes, providing a new perspective on classic adaptive hypotheses concerning this fundamental shift in vertebrate biodiversity.