Considerable variation is inherent both within and between comparative physiological data sets. Known sources for such variation include diet, gender, time of day and season of experiment, among many other factors, but a meta-analysis of physiological studies shows that surprisingly few studies report controlling for these factors. In fact, less than 3% of comparative physiological papers mention epigenetics. However, our understanding of epigenetic influences on physiological processes is growing rapidly, and it is highly likely that epigenetic phenomena are an additional 'hidden' source of variation, particularly in wild-caught specimens. Recent studies have shown epigenetic inheritance of commonly studied traits such as metabolic rate (water fleas Daphnia magna; emu, Dromaius novaellandiae), hypoxic tolerance, cardiac performance (zebrafish, Danio rerio), as well as numerous morphological effects. The ecological and evolutionary significance of such epigenetic inheritance is discussed in a comparative physiological context. Finally, against this context of epigenetic inheritance of phenotype, this essay also provides a number of caveats and warnings regarding the interpretation of transgenerational phenotype modification as a true epigenetic phenomenon. Parental effects, sperm storage, multiple paternity and direct gamete exposure can all be confounding factors. Epigenetic inheritance may best be studied in animal models that can be maintained in the laboratory over multiple generations, to yield parental stock that themselves are free of epigenetic effects from the historical experiences of their parents.