Abstract The objective was to compare taste-and-spit pleasantness ratings of open sandwiches to ratings after ad libitum consumption. In the latter test, open sandwiches were ingested one at a time at a laboratory breakfast. Three rye breads of different acid and NaCl concentrations (non-sour + normal NaCl, sour + low NaCl, sour + normal NaCl) were evaluated with butter or margarine with two levels of NaCl (1 or 2 per cent). Subjects ( N = 27) rated the samples more pleasant after consumption than in taste-and-spit tests. Mean pleasantness ratings by sample in taste-and-spit and postconsumption tests correlated with the amount of bread consumed ( r = 0.63, r = 0.82, respectively). The taste-and-spit ratings of individual subjects correlated poorly with the amount of bread consumed, only three correlation coefficients out of 27 being significant. Taste-and-spit pleasantness ratings can perhaps be used to predict the average consumption of a product but not consumption by individual subjects. The laboratory tests should be developed further to reflect better food acceptability in natural conditions.