In this investigation, a case study approach was used with two preschool children with language delays to determine whether videotaped self-modeling (VSM) intervention would influence their expressive language development. Language samples of both children were videotaped and then edited to leave only the best examples of the target language skills for each child. The children then watched the videotapes. Results indicated negligible gains for one participant and significant gains in mean length of utterance and the amount of distinguishable utterances by the other child as a result of the VSM intervention. Discussion focuses on differences between the two children--one was "distractible, imitative, but with no gains" and the other was "attentive, no verbal imitation, and positive gains." Results raise questions about the role of verbal imitations in learning as well as the role of attention difficulties and developmental readiness. Study details, letters, and forms are attached. (Contains 20 references.) (DB)