In this review of the Dark Emu debate we start out by summarising Bruce Pascoe's original work and Peter Sutton and Keryn Walshe's critique. However, the majority of this contribution is to place this Australian‐focussed debate into broader conceptual, methodological and evidential contexts generally associated with the investigation of early agriculture in other parts of the world. If we are to apply the term “agriculture” to Aboriginal plant management practices, then this requires a global, rather than a continental‐centric comparative perspective. We argue debates regarding the character of plant exploitation practices on the Australian mainland, including whether they included forms of agriculture or experimental horticulture, have been hindered by a lack of terminological clarity, the absence of a methodological framework to assess empirically verifiable evidence, and – even more problematically – a lack of relevant data on the putative plants and practices involved. Here, terminology is clarified and a bottom‐up, practice‐based method is advocated for the assessment of recent (using oral, visual and written histories) and ancient (using archaeological, archaeobotanical and palaeoecological evidence) forms of food plant exploitation in Australia. The terminology and methodological framework are heuristically applied to three scenarios: (1) ethnographic and historical records for the exploitation of underground storage organs (USOs) on the Australian mainland; (2) historical documentation regarding the botany, potential human roles in dispersal, and Aboriginal exploitation of banana (Musa spp.), taro (Colocasia esculenta) and greater yam (Dioscorea alata) in northern Australia and (3) archaeobotanical evidence for the exploitation of USOs and other plants from The Top End.