Nearly all eukaryotes possess a mitochondrial (mt) genome. Mt genes are popular targets for phylogenetic and population genetic analyses in animals due to their maternal inheritance, high mutation rates, and ease of amplification with ‘universal’ PCR primers. Variation in mt sequences is often assumed to be neutral due to the extreme functional constraints on oxidative phosphorylation (OXPHOS). However, the wealth of mt genetic data from animals is now being used to turn this assumption on its head: positive selection on mt genes has been described in diverse animal lineages. While the idea of strict mt neutrality has been systematically obliterated, cases of putative positive selection on mt genes often lack follow up experiments examining the fitness effects of mt variation. Evidence for positive selection in such studies often comes from elevated d_N/d_S ratios – the ratio of nonsynonymous to synonymous substitution rates. However, relaxed purifying selection can also produce this pattern. Here, we examined mt genomic data from several animal lineages where positive selection on mt genes has been suggested. We used tools that specifically disentangle signatures of positive selection vs. relaxed purifying selection. We find evidence in some lineages for significant, notable relaxed selection on mt genes. This challenges previous assumptions that purifying selection is unwavering in animal mt genomes. We discuss follow-up experiments underway to test these results by examining mt function in diverse lineages.