We present a genomic data set comprised of the coding DNA sequences of 5162 loci from 90 vertebrate species, including 82 mammals. The loci were aligned with their protein sequences. The aligned protein sequences were then back translated into their original DNA sequences. The alignments were further filtered to remove individual sequences from each alignment exhibiting long branches or other unusual features. The data is deposited in figshare (http://figshare.com/articles/cds\_5162.zip/6031190) and will be useful as a test data set for large-scale phylogenomic analysis.