The present study examined the effects of repeated midmorning consumption of novel-flavoured low- and high-energy yoghurt drinks on subsequent energy intake at lunch in 69 adults under actual use conditions. Subjects consumed 200 ml of low- and high-energy yoghurt drinks (67 and 273 kcal/200 ml, respectively), with 20 exposures to each drink on alternate days. Analyses focused on the development of compensation for the differences in energy content of the beverages, due to learned satiety. Results revealed incomplete energy compensation for the beverages, both at first exposure and also after 20 exposures. Relative to the no-yoghurt condition, energy intake compensation (+/-S.E.M.) averaged 39+/-36% for the low-energy yoghurt and 17+/-9% for the high-energy version, with no evidence of any change in compensation with repeated exposures. When the flavours of the yoghurt drinks were covertly switched after 20 exposures, subjects increased their energy intake after the high-energy yoghurt drink containing the flavour that was previously coupled with the low-energy yoghurt drink. Vice versa, however, when subjects switched to the low-energy yoghurt drink containing the high-energy flavour, subjects ignored the flavour cue and ate the same lunch size regardless of the energy in the yoghurt drink. We conclude that adults do not readily acquire accurate conditioned adjustments for the energy content in food after repeated experience with the food in free-living natural-eating conditions.