Convergent evolution of complex phenotypes like eyes offers us an important opportunity to investigate the redeployment and divergence of genetic components used to build those complex phenotypes. While eyes evolved convergently among animals with sophisticated neural machinery to process visual information, they surprisingly also evolved in animals with simpler nervous systems and highly capable of eye regeneration, such as jellyfish. Because eyes are formed by discrete parts, each with known genetic pathways in model systems, they are an ideal system for understanding the evolutionary trajectories underlying convergent evolution of complex phenotypes. Did eyes in jellyfish evolve convergently? And if so, to what extent do they employ similar genes and where did their parts come from? I explored these questions by integrating evidence at several levels of biological organization. First, I inferred the largest cnidarian species phylogeny to date, which allowed me to test how many times jellyfish eyes evolved. I found that eyes originated convergently at least eight times among the swimming jellyfish stage. Next, I focused on three species with convergent eyes to investigate the extent to which vision genes differentially expressed in their eye-bearing tissues were similar. I found that most genes involved in eye development and phototransduction pathways in convergent eyes are upregulated in a lineage-specific way. Comparing these findings with previous knowledge from a few other jellyfish species suggested that eyes belonging to the same evolutionary origin might employ a very predictable set of vision genes. Finally, I asked what the ancestral organismal function of photoreceptor cells was before they became part of jellyfish eyes. Based on experiments in four species of cnidarians, I found support for the hypothesis that ancestral photoreceptor cells modulated the discharge of stinging cells in the cnidarian ancestor. Such an ancient photosensory function could have long sustained a role for photoreceptor cells and perhaps facilitated multiple eye origins in the group. My dissertation work advanced a new emergent system for comparative research on eye evolution, development and function, underscored the potential for convergent eyes to have mostly unique evolutionary trajectories, and uncovered ancient sensory tasks that predated eye origins.