A fundamental question in evolutionary biology is the extent to which convergent changes in phenotype are driven by convergent changes at the sequence level. Recent phylogenetic work strongly supports multiple independent losses of flight in the ratites (ostriches, rheas, kiwis, cassowaries, and emu), providing a compelling system to understand the degree of genomic convergence associated with flight loss. We have sequenced 10 new high-quality palaeognath genomes (including 7 flightless ratites and 3 volant tinamous), and aligned these with 32 existing sequenced genomes of reptiles and birds to identify 284,000 conserved non-exonic elements (CNEEs) greater than 50 bp in length. Of these, 15,000 are specifically accelerated and/or lost in at least one ratite lineage. We show that a disproportionate number of these elements are accelerated in multiple independent ratite lineages, suggesting that convergent changes in putative regulatory sequences may be associated with independent losses of flight. Genes nearest to convergently accelerated CNEEs are enriched for developmental functions, and include several candidate genes with roles in limb development. Limb development is particularly relevant to loss of flight as forelimbs (wings) are reduced in size or lost in all ratites. Ongoing experimental work aims to determine whether genes close to convergently accelerated CNEEs are differentially expressed during the development of volant (chicken) and flightless (emu) species, and ultimately whether the sequence changes occurring in convergently accelerated CNEEs are associated with detectable changes in regulatory function. Overall, our results suggest that convergent evolution of non-coding regulatory sequence is an important driver of convergent phenotypic evolution in flightless ratites.