The effects of two introductory styles on substantive information provided by Japanese children were investigated. Children aged 5–6 years (n = 42) and 7–8 years (n = 40) took a simulated physical examination. One week later, the participants were interviewed about the examination in a narrative introductory condition using open-ended questions, or a non-narrative introductory condition using directive and yes/no questions. Substantive information provided by the children in the two conditions was compared. Results indicated that in the narrative condition, the participants provided more accurate information about the examination in response to open-ended questions. Additionally, the 7–8 year-old children in the narrative condition provided a wider range of critical information. These findings suggest that an open-ended introductory style is effective in eliciting more accurate information from children, including preschoolers, and more varied information from school-age children.