The roots of modern evo-devo can be traced back to the comparative anatomy of the 19th century. Inheriting from this tradition, the field has maintained a mechanistic approach to understanding the origins of distinct animal morphologies. While this focus has produced a valuable body of work, we argue here that a fuller understanding of why species diverge morphologically must be centered on the selective forces driving divergence, and these forces ultimately reside in the ecological context in which organisms live and reproduce. We discuss reasons why we expect many morphological novelties to evolve largely secondarily to, and often as a by-product of, primary selection on life-history traits. By shifting the focus to proximate evolutionary causes, our perspective necessarily prioritises selection experiments as a means of empirical testing. We outline experimental approaches designed to dissect the role of ecological variables in the evolution of animal development and morphology, and we show how methods and advances in fields as diverse as population genomics and ecological stoichiometry can contribute to progress in this direction. J. Exp. Zool. (Mol. Dev. Evol.) 324B: 383–392, 2015. © 2014 Wiley Periodicals, Inc.