It has been well established that chronic stress has a negative impact on the welfare of animals under our care. From the stereotypic behaviors observed in high numbers in zoo and laboratory animals to the decreased milk let-down observed in roughly handled dairy cows the effect of housing and handling practices have been demonstrated clearly by the animals in our care. Although the impact of rough handling and barren environments have been studied many of the more subtle effects that are unique to animal science, especially non-traditional laboratory species, have not been examined as closely. The objective for this paper is to review the effect of common data collection systems and environments on management, behavior, and welfare of agriculture research animals and through this introduce a source of variation. The 3 R’s of ethical animal science—Replace, Reduce, and Refine—are used to minimize the impact of research on animals and to reduce the number of animals used in research. Animal behavior and the associated welfare consequences may be sources of variation in many research studies. As an example, to collect physiological measures it is often necessary for animals to be housed in systems that vary greatly from non-research animals. These different housing and handling methods are not studied as extensively in animal agriculture and may be having more of a negative effect on welfare than expected. For this reason, it is important that we understand the effect that different research practices have on animals and accompanying research results. Many of our agriculture species are social animals that have evolved as a prey species. However, many studies require isolation and extensive human observation. Both of which alter animal behavior and welfare. Alternatively, measurement devices, such as feeding gates, may affect normal behavior. Differences in social dominance, prior experiences, and temperament can alter the ability of animals to learn to interact with these systems. For these reasons, an understanding of animal behavior is essential to minimize welfare effects and the accompanying variation in study results.