Rats were trained to avoid a sugar solution through pairing with LiCl toxicosis (upper gastrointestinal tract discomfort), shock (peripheral pain), or high levels of lactose (lower gastrointestinal tract discomfort). Among animals matched for strength of avoidance of the sugar solution, only the LiCl group showed orofacial responses (e.g., gaping) indicative of distaste; the other groups continued to show positive orofacial responses to the sugar solution. These results, in conjunction with recent results on humans, are interpreted to represent a distinction between food rejection based primarily on unpalatability (distaste) and food rejection based primarily on anticipated negative consequences of ingestion (danger). The results also support the hypothesis that upper gastrointestinal distress (most probably nausea) plays a special role in negative palatability shifts (acquired distastes). These results have implications for the understanding of predispositions in learning and suggest important differences in the quality (readout) of different types of associations. Prior research, by relying on intake measures alone, was insensitive to the differences revealed here by monitoring a wider range of responses.