We classify the explanation of the base rate fallacy into three positions: (1) the neglect view, which argues that the neglect of base rates causes the fallacy; (2) the frequency view, which claims that the fallacy will disappear if information is presented in the natural frequency format instead of the probability format; and (3) a family of theories that focus on mental representations of the task structure. In this paper, three experiments examined the validity of the equiprobability hypothesis, which is one of the theories that emphasize mental representations of the task structure. In Experiments 1 and 2, although we used tasks in which the neglect view predicted high performances, correct answers remained infrequent. In Experiment 3, facilitation by the natural frequency format could barely be distinguished from the effect predicted by the equiprobability hypothesis. Consequently, we suggested that the neglect view was inappropriate and the frequency view could be reinterpreted according to the equiprobability hypothesis.