Researchers recommend collecting repeated or prolonged natural language samples to supplement direct assessments and anecdotal reports in language studies for autistic children. Several studies have used the Language Environment Analysis® (LENA) system to collect language samples from autistic children. However, there has been little research that investigates the reliability of using the LENA system for autistic children in an educational setting such as a classroom. The current study compared language data disaggregated by the LENA system from 40 autistic children in educational settings with data from human coders. Specifically, we calculated three separate correlational and reliability analyses between the LENA system and human coders. Results showed that although the aggregated coefficients could be interpreted as fair, LENA systems should be used with caution due to high variability between the LENA system and human coders. Implications for future research and limitations are also discussed.