The present article reviews rat experiments using aversive stimuli in which some element of uncertainty was involved. Nine cases of conditions that produce uncertainty are described, and the relevant experimental facts are examined. Rats were found to prefer situations involving certainty to those involving uncertainty, and the rats' basal rate of responding was found to be less in suppression situations involving uncertainty about aversive environmental events than in those involving relative certainty. These results were obtained when the aversive stimuli the rats received were physically matched in quantity among the conditions compared. The results are discussed with reference to five hypotheses. It is concluded that the function of the concept of uncertainty is to organize a class of studies that are otherwise unrelated to each other.