The principles and tenets of management require action to avoid sustained abnormal/pathological conditions. For the sustainability of interactive systems, each system should fall within its normal range of natural variation. This applies to individuals (as for fevers and hypertension, in medicine), populations (e.g. outbreaks of crop pests in agriculture), species (e.g. the rarity of endangerment in conservation) and ecosystems (e.g. abnormally low productivity or diversity in ‘ecosystem–based management’). In this paper, we report tests of the hypothesis that the human species is ecologically normal. We reject the hypothesis for almost all of the cases we tested. Our species rarely falls within statistical confidence limits that envelop the central tendencies in variation among other species. For example, our population size, CO2 production, energy use, biomass consumption and geographical range size differ from those of other species by orders of magnitude. We argue that other measures should be tested in a similar fashion to assess the prevalence of such differences and their practical implications.