Children at ages 4 and 7 were presented three stories, one in each of three causality versions, in order to determine whether inference ability and recall are influenced by the strength of the causal chain in the story. For each story, children were asked two inference questions about logical causality, two questions constrained by story information, and two unconstrained inference questions. Three question-timing conditions were used to address issues concerning when inferences are drawn in story comprehension and how they are influenced by the total amount of information provided. Questions were asked either on-line, or at the end of the story, or not at all. Following completion of each story, children were asked to retell it, and were then asked premise information questions.