Abstract Conodont elements have high rates of morphological evolution, but the drivers of this disparity are debated. Positive allometric relationships between dimensions of food-processing surfaces and entire P 1 elements have been used to argue that these elements performed mechanical digestion. If involved in food processing, the surface of the element should grow at a rate proportional to the increase in energy requirements of the animal. This inference of function relies on the assumption that the energy requirements of the animal grew faster (≅ mass 0.75 ) than the tooth area (≅ mass 0.67 ). We reevaluate this assumption based on metabolic rates across animals and calculate the allometry in platform-bearing P 1 elements of Late Triassic co-occurring taxa, Metapolygnathus communisti and Epigondolella rigoi , using 3D models of ontogenetic series. Positive allometry is found in platform and element dimensions in both species, supporting a grasping-tooth hypothesis, based on the assumption that metabolic rate in conodonts scaled with body mass similarly to that in fish and ectotherms. We also calculate the curvature of the P 1 platform surface using the Dirichlet normal energy (DNE) as a proxy for diet. DNE values increase with body mass, supporting the assumption that conodont metabolic rates increased faster than mass 0.67 . We finally find that adults in both taxa differ in their food bases, which supports trophic diversification as an important driver of the remarkable disparity of conodont elements.