Previous research suggests that increasing beverage protein content enhances subsequent satiety, but whether this effect is entirely attributable to post-ingestive effects of protein or is partly caused by the distinct sensory characteristics imparted by the presence of protein remains unclear. To try and discriminate nutritive from sensory effects of added protein, we contrasted effects of three higher-energy (about 1·2 MJ) and one lower-energy (LE: 0·35 MJ) drink preloads on subsequent appetite and lunch intake. Two higher-energy drinks had 44 % of energy from protein, one with the sensory characteristics of a juice drink (HP − , low-sensory protein) and the second a thicker and creamier (HP+, high-sensory protein) drink. The high-carbohydrate preload (HC+, high-sensory carbohydrate) was matched for thickness and creaminess to the HP+ drink. Participants (healthy male volunteers, n 26) consumed significantly less at lunch after the HP+(566 g) and HC+(572 g) than after HP −  (623 g) and LE (668 g) drinks, although the compensation for drink energy accounted for only 50 % of extra energy at best. Appetite ratings indicated that participants felt significantly less hungry and more full immediately before lunch in HP+ and HC+ groups compared with LE, with HP −  being intermediate. The finding that protein generated stronger satiety in the context of a thicker creamier drink (HP+ but not HP − ) and that an isoenergetic carbohydrate drink (HC+), matched in thickness and creaminess to the HP+ drink, generated the same pattern of satiety as HP+, both suggest an important role for these sensory cues in the development of protein-based satiety.