SUMMARY. An experiment was carried out to investigate 7-8 year-olds' use of implicit inferences in understanding stories. Two groups of children, differentiated by their ability at text comprehension, read four short stories and were asked a series of questions after each one. The results showed that skilled readers were better than less skilled readers at answering questions from memory shortly after reading a story, both when the questions could be answered directly from the text, and when they required an inference. However, when the text was made available, the less skilled group remained poorer at answering questions that required an inference, although their performance on literal questions improved to the same level as that of the skilled group. Thus, the experiment supports the idea that a major distinguishing characteristic of skilled readers is that they are good at making inferences, which enable them to relate the ideas in a text one to another and to general knowledge. Additionally, the results provide no support for the claim that differences in ability to make inferences can be attributed to differences in memory for prose: the less skilled group were poorer at making inferences even when the text was available to them.INTRODUCTION IN order to understand stories adequately, children must be able to draw appropriate inferences spontaneously when they hear or read a story. Several studies have shown