Two experiments were conducted to examine whether children would take into account the foreseeability of the outcomes an actor would accidentally produce in judging responsibility. The foreseeability was expected to differ according to situations connected with the event or according to the ability of the actor to foresee the effect of his conducts. Second, fourth and sixth grade children and undergraduates participated in the study. In Experiment 1, subjects were presented 10 stories in which an actor uninten -tionally caused a negative outcome. Subjects were asked to rate the degree of respon -sibility and foreseeability attributed to the actor. In Experiment 2, subjects were presented 5 pairs of stories. In each pair two actors of different age or different occupation were to cause the identical outcome. Subjects were asked to judge which person was more responsible and were asked to give reasons proving their judgements. Subjects of all age groups used the foreseeability of the situation in judgments. However, the 2nd graders and 4th graders would not infer the difference in actors' ability to foresee the outcomes when the difference was not explicit.