In this study, we used bioinformatic approach to analyze the compositional features and codon usage bias (CUB) of ATP6 and ATP8 genes among three groups, namely, fishes, aves, and mammals which thrive in three different habitats as no work was reported yet. The coding sequences of these genes were retrieved from the National Center for Biotechnology Information to explore the similarities and dissimilarities of codon usage of each gene among these groups. Low values of synonymous codon usage order in fishes, aves, and mammals for ATP6 and ATP8 genes suggested that the CUB of ATP6 and ATP8 genes was low. In ATP6 gene, overall GC contents in fishes, aves and mammals were (mean ± SD) 44.09 ± 3.10, 46.65 ± 1.90, and 39.41 ± 2.89%, respectively, whereas in ATP8 gene, the overall GC contents were 42.76 ± 4.38, 44.16 ± 2.43, and 34.19 ± 3.82% in fishes, aves, and mammals, that is, both genes were found to be AT rich. In ATP6 gene, the codon AGC was overrepresented in fishes but under‐represented in aves and mammals, whereas in ATP8 gene, the codon GCC was overrepresented in fishes but underrepresented in aves and mammals. The pattern of codon usage was different in these genes and varied among groups as evident from correspondence analysis. The slope of the regression line in neutrality plot was lower than 0.5, which revealed that the role of natural selection was higher than mutation pressure in shaping the CUB in ATP6 and ATP8 genes.