In this paper we assess the potential for developing mimics in situations where detailed data on the ecophysiology and ecosystem function of the native ecosystem are not available. We examine two ecosystems in south western Australia, and draw together available information on these. We discuss woodlands dominated by Eucalyptus salmonophloia, which formerly covered much of the valley bottom land in the Western Australian wheatbelt, and kwongan or heath which occurs on sandplain soils at the top of the landscape. From both of these we derive possible mimics which utilise aspects of the native vegetation which we hypothesise contributed to ecosystem function and resilience, but which still allow the production of a harvestable agricultural product. We conclude that situations for which we have adequate ecological knowledge from which to develop sound mimics are likely to be rare. We hence have to be prepared to make best-bet estimates and use these to develop experimental mimic systems.