This study examined acquired liking of flavour preferences through flavour-flavour and flavour-nutrient learning under hungry or sated conditions in a naturalistic setting. Each participant consumed one of three versions of a test drink at home either before lunch or after lunch: minimally sweetened (3% sucrose, 40 kcal), artificially sweetened (3% sucrose 40 kcal plus artificial sweeteners ASPARTAME) and sucrose-sweetened (SUCROSE: 9.9% sugar, 132 kcal). The test drink was an uncarbonated peach-flavoured iced tea served in visually identical drink cans (330 ml). Participants preselected as "sweet likers" evaluated the minimally sweetened flavoured drink (conditioned stimulus, CS) in the same state (hungry or sated) in which they consumed the test drink at home. Overall, liking for the CS flavour increased in participants who consumed the SUCROSE drink, however, this increase in liking was significantly larger when tested and trained hungry than sated, consistent with a flavour-nutrient model. Overall increases in pleasantness for the CS flavour in participants who consumed the SUCROSE drink when sated or the ASPARTAME drink independent of hunger state, suggest that flavour-flavour learning also occurred. These results are discussed in light of current learning models of flavour preference.