Individuals with high functioning autism present with discrepancies in communication and language skills as compared to typical age-matched cohorts. Formal testing often identifies children with ASD as having greater deficits in receptive language than expressive language skills. However, expressive language is frequently filled with overlearned language and repetition. This masks limitations in linguistic performance at the discourse level while allowing for achievement of scores within normal limits on standardized tests of expressive language competence. These tests give an inaccurate picture of novel linguistic abilities related to flexibility and creativity in discourse. The aim of this study is to compare expressive language skills in children with autism (n=4) to normal controls (n=4) using two subtests of expressive language competence from the CELF-5 and a formal measure of expressive language performance (TNL). It was hypothesized that children with ASD would perform significantly lower on formal measures of expressive language performance than measures of expressive language competence as compared to normal controls. Results supported an interaction between groups and test scores. The subjects with ASD performed significantly lower on the test of expressive language performance as compared to normal controls. Scores by the ASD group on subtests of language competence were equivalent or better than normal controls. These findings suggest the use of multiple formal measures in assessing expressive language at the discourse level in the ASD population. This will increase eligibility and improve goals for targeting novel language use during creative expression.