We typically think of morphological structures or traits as evolved for a single function, but the natural world is rarely that simple. More often, the parts of an organism contribute to the performance of multiple behaviors across a range of contexts. How this multifunctionality affects the evolution of a trait is an open question in biology. On one hand, a greater number of functions might constrain evolution due to an increased number of trade-offs – i.e. if each function selects for different trait optima, few phenotypic states can perform all functions. Alternatively, a greater number of functions may facilitate diversification by adding complexity to the adaptive landscape of a trait. To effectively test between these two hypotheses requires both a suitable study system and a dataset of considerable taxonomic breadth and depth. Therefore, we used geometric morphometrics to characterize the wings of 2,332 individual birds from 959 species, dwarfing previous studies of avian wing shape. We then tested whether the rate of evolution on the avian wing is determined by the number of fluids in which that wing is used for locomotion. Importantly, five lineages of birds use their wings for locomotion in both air and water, while another five lineages of equally volant birds rely on their feet for aquatic locomotion. Thus, this system offers an exciting opportunity to explore how multifunctionality influences the evolutionary process, while controlling for ecology. After accounting for variation due to wing size, we find strong evidence that multifunctionality constrains the rate of wing shape evolution. In the coming months, we will further explore this dataset using multivariate statistical methods to test whether intraspecific variation and evolutionary mode also vary predictably with the number of wing functions.