Flavor-flavor and flavor-nutrient associations can modify liking for a flavor CS, while flavor-flavor associations can also modify the sensory experience of the trained flavor. Less is known about how these associations modify behavioral responses to the trained CS. To test this, 60 participants classified as sweet likers were divided into five training conditions with a novel flavor CS. In the flavor-flavor only condition, participants consumed the target CS in a sweetened, low-energy form, with energy (maltodextrin) but no sweetness added in the flavor-nutrient only condition and both energy and sweetness (sucrose) in the combined flavor-flavor, flavor-nutrient condition. Comparison groups controlled for exposure to the CS, and repeat testing. Training was conducted in a hungry state on four non-consecutive days. To test for acquired changes in evaluation and intake, the flavor CS was processed into a low-energy sorbet, which was evaluated and consumed ad libitum on test days before and after training. Liking for the flavor CS increased only in the sucrose-sweetened condition, but intake increased significantly in both high-energy conditions. In contrast, rated sweetness of the sorbet increased in both sucrose-sweetened and aspartame-sweetened conditions. These findings suggest that liking changes were maximal when flavor-flavor and flavor-nutrient associations co-occurred, but that behavioral changes were specific to flavor-nutrient associations.