This study focuses on the relationship between teacher-text-child interactions during shared picturebook reading and its effect on children's narrative comprehension. Using a micro-analytic approach, we studied nine large-group bookreading sessions of one experienced teacher and her classroom of 17 preschoolers from low-income backgrounds over a school year. We coded the bookreading interactions according to the level of analytic rigor demanded by the questions the teacher asked (literal, simple inferential, complex inferential), the children's responses to these questions (incorrect, adequate, correct), and the teacher's follow-up responses (evaluative, elaborative), including additional follow-up questions and prompts elicited by the children's responses. We examined whether picturebook difficulty (low, medium, high) was related to the demand level of the teacher's questions and the children's responses. We also examined whether children's participation in these interactions was associated with children's narrative comprehension, which was tested at the beginning and end of the school year. Overall, the teacher consistently asked a higher proportion of inferential (80%) than literal questions (20%). Medium-difficulty level books elicited more complex inferential questions than easier or harder books. Children correctly answered such questions from medium-difficulty books more often than easy or difficult ones. Increases in children's narrative comprehension from pretest to posttests were related to the children's frequency of active participation and the accuracy of their responses. These results have implications for how picturebook choice and type of discussion can promote narrative comprehension among preschoolers from low-income backgrounds.