This paper examines the effects of two different types of treatment provided for young children with a specific language impairment (SLI). Two groups of seven children matched for language skills were randomly allocated to one of two treatment conditions. In the first, the children were given group treatment three times a week over a seven week period. At the same time the parents were provided with weekly sessions encouraging them to focus on specific aspects of their child's interaction and language development. In the second condition, children were provided with group treatment equivalent to that in the first condition but parents were not provided with the additional support. Change was measured using video analysis of parent/child interaction in a fixed context before and after the intervention. As predicted, child initiations increased following both groups providing input to children. The parents who were involved in the intervention increased their use of specific referent terms directly relevant to the nonverbal context. However, very little difference was noted in the behaviour of parents who were not involved in intervention. The clincial implications of these findings are discussed.