The concept of the adaptive landscape has been an invaluable tool to the field of modern evolutionary biology by providing a representation of how fitness and selection vary within populations of organisms. Although originally implemented in the study of population genetics and microevolution, paleontologist G.G. Simpson expanded the idea to explain phenotypic change over macroevolutionary timescales. Despite decades passed, the practical application of Simpson’s adaptive landscape in macroevolution has been limited by computational challenges in quantifying performance outcomes of morphological traits and how they vary across a vast phenotypic landscape. Here I will introduce a method called “functional adaptive landscape analysis” which determines the functional trade-offs associated with morphological adaptation to differing selection regimes and I will then apply this method to explore two major transitions in vertebrate evolution. First, I will tackle the fish-to-tetrapod transition and the origin of land locomotion by modelling the evolution of the humerus (upper arm) bone. I will show that the earliest tetrapods occupied a performance valley between water and land adaptive peaks, but that adaptations suggest some capacity for terrestrial locomotion. Second, I will discuss the ‘reptile-to-mammal’ transition and the evolution of the mammalian backbone. Using vertebral morphometric data and experimental biomechanics, I will test the lateral-sagittal functional paradigm and demonstrate how this long-held idea is too simplistic to explain mammalian backbone evolution. Together, I hope to establish the utility of functional adaptive landscapes in quantitatively testing longstanding questions in the vertebrate fossil record and its potential application to unravelling the relationship between form, function, and adaptation across deep time.