Purpose This study evaluates the effectiveness of a nonword repetition (NWR) task in discriminating between Palestinian Arabic-speaking children with developmental language disorder (DLD) and age-matched typically developing (TD) children. Method Participants were 30 children with DLD aged between 4;0 and 6;10 (years;months) and 60 TD children aged between 4;0 and 6;8 matched on chronological age. The Arabic version of a Quasi-Universal NWR task was administered. The task comprises 30 nonwords that vary in length, presence of consonant clusters (CCs) and wordlikeness ratings. Responses were scored using an item-level scoring method to assess the diagnostic accuracy of the task. Receiver operating characteristic curve analysis was conducted to determine the best cutoff point with the highest sensitivity and specificity values, and likelihood ratios were calculated. Results Children with DLD scored significantly lower on the NWR task than their age-matched TD peers. Only the DLD group was influenced by the phonological complexity of the nonwords, with nonwords with two CC being more difficult than nonwords with no or only one CC. For both groups, three-syllable nonwords were repeated less accurately than two- and one-syllable nonwords. Also, high word-like nonwords were repeated more accurately than nonwords with low wordlikeness ratings. The best cutoff score had sensitivity and specificity of 93% and highly informative likelihood ratios. Conclusions NWR was an area of difficulty for Palestinian Arabic-speaking children with DLD. NWR showed excellent discriminatory power in differentiating Arabic-speaking children diagnosed with DLD from their age-matched TD peers. NWR appears to hold promise for clinical use as it is a useful indicator of DLD in Arabic. These results need to be further validated using population-based studies. Supplemental Material https://doi.org/10.23641/asha.14880360.