This experiment investigated performance on a reasoning task in which the subjects had to say which of four half-hidden cards they needed to see fully in order to determine the truth or falsity of a conditional sentence. In one experimental condition examples of each of the four types of card were fully displayed to remove any difficulty due to limitation of memory or imagery. No more correct solutions were obtained under this condition than under a control condition in which no such aid was given. After they had made their selections all subjects wrote down their reasons for them. Analysis of the different selections, and the reasons given for them, corroborated a previously proposed information-processing model, postulated to explain the cognitive processes involved in tasks of this type.