There is now convincing evidence that many inflammatory diseases are precipitated, or at least exacerbated, by unfavourable interactions of the host with the resident microbiota. The role of gut microbiota in the genesis and progression of diseases such as inflammatory bowel disease, obesity, metabolic syndrome and diabetes have been studied both in human and in animal, mainly rodent, models of disease. The intrinsic variation in microbiota composition, both within one host over time and within a group of similarly treated hosts, presents particular challenges in experimental design. This review highlights factors that need to be taken into consideration when designing animal trials to investigate the gastrointestinal tract microbiota in the context of inflammation studies. These include the origin and history of the animals, the husbandry of the animals before and during experiments, details of sampling, sample processing, sequence data acquisition and bioinformatic analysis. Because of the intrinsic variability in microbiota composition, it is likely that the number of animals required to allow meaningful statistical comparisons across groups will be higher than researchers have generally used for purely immune-based analyses.