While numerous qualitative social scientific analyses of (environmental) epigenetics have been published, we still lack a macro-level, quantitative assessment of the field of epigenetics as a whole. This article is aimed at filling this gap. Mobilizing an extended version of the Web of Science, we constituted a corpus of 199,484 documents (articles, reviews, editorial material, etc.) published between 1991 and 2017 and performed several scientometric analyses to map out the development and structure of the epigenetics field. Three main results were drawn from these investigations. First, contradicting the hope expressed by some social scientists that their disciplines will find solace in epigenetics’ social biology, it is striking that the scientists, journals and institutions that drive most of the research in the field are overall little concerned with social and environmental dimensions of gene expression. Second, and confirming existing qualitative analyses, we find that epigenetics is constituted by diverse networks of scholars, institutions and research specialties that enjoy relative autonomy from each other and approach epigenetics through different thematic interests, from cognitive functions to cancer, to DNA methylation in plants and molecular biology. Third, findings obtained from the bibliographic coupling showed that these different networks became more and more autonomous over the last decade, which suggests that we are currently witnessing the constitution of a scientific archipelago akin to that of behavior genetics (Panofsky, 2014: 33) rather than to a discipline per se. At the same time, this differentiation was less pronounced conceptually speaking, as we also observed a clear standardization of the keywords used in epigenetics articles between 1991 and 2017, with DNA methylation and RNAs serving as rallying signs for different communities of researchers.