Purpose: The objective of this study was to examine the effects of coaching by speech-language pathologists on educators' interactive shared book reading, children's participation in shared reading, and children's language development. Method: Thirty-two educators and small groups of preschoolers were randomly assigned to experimental and comparison groups. The experimental group (n = 15) received 4 in-service workshops plus 5 individualized coaching sessions. The comparison group received only the 4 workshops. Participants were video-recorded during a shared book reading activity with a small group of children at pretest and posttest. The video recordings were transcribed and coded to yield measures of conversations, educators' questions, and children's responses. The mean length of utterances of the children's responses was also calculated. Results: There were no significant Time x Group interaction effects for the number and length of shared reading conversations or for the number of participants in these conversations. However, significant Time x Group interactions were observed for the use of educators' experiential reasoning questions, children's experiential reasoning responses, and the mean length of utterances of children's responses. Conclusion: These results suggest that coaching increases educators' use of inferential questions, enhancing an interactive shared-reading strategy that had a direct impact on the children's quality and complexity of language. (author abstract)