This investigation included two phases of inquiry that examined the effects of peer modeling upon the play scripts of children with specific language impairment (SLI). The first study employed a pretest-posttest control group design involving two groups of children with SLI (10 who participated in the experimental treatment and 10 controls) and a group of peer models (10 children with normal language development). The treatment involved dyadic play sessions in which children with SLI were paired with a normal language peer model. Significant differences were found between the play script reports of the experimental (SLI-E) and control groups (SLI-C) of children with specific language impairment. The second study, utilizing single-case methodology, involved 6 children with SLI who participated in the control group of Study 1, plus 2 peer models. Play dyads consisted of either two children with SLI or one child with SLI and a normal language peer. Results of this study provided support for the contention that play interactions with normal language peers facilitates increases in the play-script reports of children with SLI.