The role of the temporal order of odor and taste was studied in two experiments, and a third experiment studied the role of odor intensity in flavor-toxicosis conditioning with thirsty rats licking water spouts in a "wind tunnel." In all experiments, odors and tastes were presented for 2 min to rats, and 30 min later, a toxin (lithium chloride) was intubated. In Experiment 1, an odor was presented 90 s before, during, or 90 s after a taste to independent groups. Experiment 2 was a within-subjects partial replication of the first. Each rat was presented with one odor, then a taste, then a second odor with each stimulus separated by 45 s. The results of Experiments 1 and 2 indicated that (a) odor alone is not associated with illness under our conditions, (b) presenting an odor and a taste at the same time potentiates the odor component so that it is associated with illness, (c) 45-s and 90-s intervals between odor and taste eliminate potentiation, and (d) taste and odor interact asymetrically; that is, odor has little affect on the development of taste-illness associations. In Experiment 3, an odor and a taste were presented simultaneously, and odor intensity varied. As odor intensity increased, the strength of the taste-potentiated odor aversion increased, whereas the aversion to the taste remained constant. However, even at the highest intensity, odor presented in the absence of taste did not result in odor aversions.