Abstract
Summary Baas‐Becking is famously attributed with the conjecture that ‘everything is everywhere, but the environment selects’. Although this aphorism is largely challenged by microbial biogeographical data, even weak versions of the claim leave unanswered the question about whether all environments that could theoretically support life contain life. In the last decade, the discovery of thermally sterilized habitable environments disconnected from inhabited regions, and habitats within organisms such as the sterile, but habitable human fetal gut, suggest the existence of a diversity of macroscopic habitable environments apparently devoid of actively metabolizing or reproducing life. Less clear is the status of such environments at the micron scale, for example, between colonies of organisms within rock interstices or on and within other substrates. I discuss recent evidence for these types of environments. These environments have practical uses in: (i) being negative controls for understanding the role of microbial processes in geochemical cycles and geological processes, (ii) yielding insights into the extent to which the biosphere extends into all spaces it theoretically can, (iii) suggesting caution in interpreting the results of life detection instrumentation, and (iv) being useful for establishing the conditions for the origin of life and its prevalence on other planetary bodies.