Conditioning involved adding one flavor (e.g., vanilla) to an artificial diet of high caloric density and a second flavor (e.g., anise) to a similar low-calorie diet. When the rats were tested with identical foods to which the flavors were added, they showed a strong preference for the high-calorie flavor. The parameters of conditioning were varied in four experiments, which showed that the conditioning of taste preferences is rapid and powerful and has generality over a range of conditioned stimulus parameters. The last experiment indicated that the unconditioned stimulus for this conditioning is probably some oral cue, such as the taste of starch, which is correlated with caloric benefit, rather than caloric benefit per se.