This study investigated the morphosyntactic abilities of Saudi Arabic-speaking children with Autism Spectrum Disorder (ASD) using a Saudi Arabic Sentence Repetition Task (Saudi-SRT) developed for this study. The first objective of the study was to determine whether Saudi Arabic-speaking children with autism living in Saudi have difficulties in morphosyntax as a function of their verbal abilities. For this reason, the participating children with autism were divided into subgroups of children with autism with normal language (ALN) and children with autism with language impairment (ALI) based on their verbal abilities. The second objective of this study was to identify whether the children’s performance differs as a function of the scoring scheme used. Towards that aim, two scoring schemes were used based on whether children repeated sentences verbatim versus whether they used the targeted structure accurately. The third objective was to address differences between the structures used in the Saudi-SRT. The study involved 62 five- to seven-year-old children who speak Saudi Arabic as their first language: 20 children with autism (n = 10 ALN and n = 10 ALI) and 42 typically developing (TD) control children. The children with autism showed difficulties in morphosyntax as a function of their verbal abilities, but the scoring scheme modulated this. In nonverbal abilities and receptive vocabulary, ALI scored lower than TD and ALN, and there was no difference between TD and ALN. As for the grammatical abilities, the less sensitive verbatim scheme showed a similar pattern in the ALN and ALI children who performed less well than the TD children; in contrast, the more sensitive structural scheme showed that in the most syntactically complex structures, the ALN children performed similarly to the TD children and better than the ALI children.