In this paper, I discuss whether in the Wason selection task, where subjects are asked to select evidence for evaluating a conditional, the most common response is indeed as irrational as many psychologists have maintained. My account of subjects’ performances focuses on (i) how the task is understood by the subjects and (ii) whether their performances can be seen as cases not so much of conditional reasoning, as of hypothesis testing. On this basis and in accordance with Grice’s account of indicative conditionals, I suppose that normally people understand the conditional as involving a “strong” connection between the antecedent and the consequent. Whenever the inference to such a connection is not blocked, subjects will inductively test whether it holds, as they do in ordinary life with reference to wider domains than the narrow one provided by the experimental context. Thus, relative to their goals, their performance could be considered as rational.