Generally fasted birds are used for estimating endogenous energy (EE) excretion, this measurement being a critical factor in determining the TME of feeds and ingredients. Evidence is presented that EE is overestimated because the fasted birds lack energy during the bioassay period and that estimation of EE should be made rather from birds in a positive energy balance. The possibility of an improved procedure for estimating EE making use of nitrogen excretion is discussed.