Two small pen trials with cattle and sheep both clearly demonstrated that while there is significant variation in residual feed intake when on high energy supply, there is negligible variation when energy supply is limited. A review of literature demonstrated that this is also the case when energy supply is limited by heat or physiological state, such as peak lactation, and in multiple species. There is little evidence of variation in efficiency of maintenance requirements, growth or lactation. Nor is there strong evidence for large variation in digestibility within breeds, despite some differences between divergent breeds. Thus, the primary source of variation in residual feed intake must be in appetite and, in variable environments, it is possible that those with greater appetite are more resilient during times of feed shortage.