Chowdhury and Ørskov (1994) observed that heat production in sheep was reduced by approximately 10% below that at fasting when the animal was offered one third of the predicted maintenance requirement through intragastric nutrition. These authors also suggested that this indicated a glucose deficient state in the fasted animal, and that as a result the heat production at fasting was artificially raised. In agreement with Ørskov and MacLeod (1990), Chowdhury and Ørskov (1994) further stated that when measuring heat production to estimate basal metabolism animals should have their heat production measured when being offered one third of maintenance rather than fasting, although this was not tested by experiment. Birnie et al., (2000) reported that fasting heat production was influenced by the condition score of the cow. The objectives of this experiment were firstly to investigate the influence of maintenance and sub-maintenance levels of feeding on the observed heat production in nonpregnant, nonlactating dairy cows offered a more normal diet and secondly to further examine the influence of cow condition score on the fasting heat production.