Hedonic ratings of foods, obtained by laboratory taste-and-spit tests, are often used to determine the optimal concentration of sugar in a commercial food. However, the predictive value of these tests on actual consumption has not been demonstrated. Does the optimal sugar concentration in a specific food, as determined by a short exposure taste test, induce maximum intake of that food? In the present experiment, a standard food (yogurt) was presented in five sucrose concentrations (from 2% to 29%) to 18 subjects. The five types of yogurts were first tasted and rated for pleasantness. Secondly, they were presented in ad lib consumption tests. Discrepancies appeared between taste test ratings and ingestive responses, e.g., in subjects who preferred medium to high sucrose concentrations, the optimal concentration (in terms of intake) was lower than the "preferred" one (as determined by taste-and-spit tests). The results of taste-and-spit should be taken cautiously when trying to predict ingestion.