There is a great deal of interest regarding the biological embedding of childhood trauma and social exposures through epigenetic mechanisms, including DNA methylation (DNAm), but a comprehensive understanding has been hindered by issues of limited reproducibility between studies. This review presents a summary of the literature on childhood trauma and DNAm, highlights issues in the field, and proposes some potential solutions. Investigations of the associations between DNAm and childhood trauma are commonly performed using candidate gene approaches, specifically involving genes related to neurological and stress pathways. Childhood trauma is defined in a wide range of ways in several societal contexts. However, although variations in DNAm are frequently found in stress-related genes, unsupervised epigenome-wide association studies (EWAS) have shown limited reproducibility both between studies and in relating these changes to exposures. The reproducibility of childhood trauma DNAm studies, and the field of social epigenetics in general, may be improved by increasing sample sizes, standardizing variables, making use of effect size thresholds, collecting longitudinal and intervention samples, appropriately accounting for known confounding factors, and applying causal analysis wherever possible, such as "two-step epigenetic Mendelian randomization."