The biota and ecological processes of the Australian arid zone remain relatively unstudied. Australian deserts are remote from population centres and are subject to unpredictable rainfall, making it difficult to plan and execute time-limited research projects. Desert research is expensive; the lack of basic infrastructure, i.e. roads, and the remoteness from major service centres means that vehicle and fuel costs are high. Increasingly, research funding is being directed to universities for postgraduate research projects, instead of more long-term, expensive projects in research centres and government departments. Postgraduate research is strictly time- and cost-limited, which can constrain research effort in desert systems where stochastic conditions prevail. However, steps can be taken to ensure that short- to medium-term desert research projects are viable and successful. Combining different approaches and methods of data collection ensure that a project can become ‘desert-proofed’ when unfavourable conditions occur. I discuss a variety of different research projects and the need to reconsider our approach to desert research, giving emphasis to the study of functions and systems instead of a single, elusive species.