Flavor preferences conditioned by glucose and fructose were compared using two training methods. With the simultaneous method preferences can be reinforced by the flavor and/or the postingestive consequences of nutrients, whereas with the delayed method preferences are reinforced only by postingestive nutritive effects. In Experiment 1, food-deprived rats displayed similar preferences for flavors (CS+) added to an 8% glucose or 8% fructose solution over flavors (CS-) added to a noncaloric saccharin solution (simultaneous conditioning). Other rats learned to prefer a CS+ flavor paired with the delayed (10 min) presentation of 8% glucose over a CS- flavor paired with delayed saccharin. Fructose failed to condition a flavor preference with the delayed paradigm. Taken together, these data suggest that the preference for a flavor mixed in a fructose solution is reinforced by the sweet taste, not the postingestive effects of the sugar. Experiment 2 tested this idea by devaluing the taste of the sugar solutions by quinine adulteration. Rats initially avoided both glucose-quinine and fructose-quinine solutions in favor of a saccharin solution. Following one-bottle training, they came to prefer the glucose-quinine but not the fructose-quinine solution over the saccharin solution. The glucose-trained rats also showed stronger preferences for sucrose-quinine solutions than did the fructose-trained rats. These findings, along with other recent data, indicate that fructose-conditioned preferences are based primarily on the sugar's palatable taste. Glucose, in contrast, can condition strong preferences based on its taste as well as its postingestive actions.