Advances in epigenetic methodologies have deepened theoretical explanations of mechanisms linking early life stress (ELS) and disease outcomes and suggest promising targets for intervention. To date, however, human studies have not capitalized on the richness of diverse animal models to derive and systematically evaluate specific and testable hypotheses. To promote cross‐species dialog and scientific advance, here we provide a classification scheme to systematically evaluate the match between characteristics of human and animal studies of ELS and DNA methylation. Three preclinical models were selected that are highly cited, and that differ in the nature and severity of the ELS manipulation as well as in the affected epigenetic loci (the licking and grooming, maternal separation, and caregiver maltreatment models). We evaluated the degree to which human studies matched these preclinical models with respect to the timing of ELS and of DNA methylation assessment, as well as the type of ELS, whether sex differences were explicitly examined, the tissue sampled, and the targeted loci. Results revealed <50% match (range of 8–83%) between preclinical models and human work on these variables. Immediate and longer‐term suggestions to improve translational specificity are offered, with the goal of accelerating scientific advance.