Purpose: To avoid misdiagnosis with developmental language disorder (DLD) in bilingual children, it is recommended to evaluate both languages. However, unlike their monolingual peers, bilingual children acquire their heritage language under adverse input conditions. Focusing on Levantine Arabic, the study evaluates the clinical utility of a standardized test for Arabic and an Arabic Language Impairment Testing in Multilingual Settings sentence repetition task (LITMUS-SRT) for ruling in/out DLD in bilingual children acquiring Arabic as a heritage language in Germany. Both tools were developed for the majority diglossic Lebanese bilingual context. However, it is still unclear whether they can be reliably applied in heritage contexts. Method: Fifty-nine children (5;6--9;1 [years;months]) participated in the study: 18 bilingual typically developing ([BiTD]) children with early exposure to German (heritage BiTD), nine heritage bilingual children with DLD (heritage BiDLD), eight typically developing late-successive bilingual children, and 24 typically developing monolingual children, who served as controls. Results: Unlike monolingual and late successive bilingual children with typical language development, many heritage BiTD children resembled their BiDLD peers, especially on measures of expressive/receptive vocabulary and morphosyntax production, which were negatively affected by early second language exposure. In contrast, the Arabic LITMUS-SRT did not disadvantage heritage BiTD children, who performed on par with their typically developing monolingual and late successive bilingual peers on most structures. BiDLD showed poor performance on all structures. Conclusions: Standardized tests assessing language domains sensitive to reduced heritage language input bear a risk of overdiagnosis with DLD in heritage contexts. The Arabic LITMUS-SRT, on the other hand, provides a fair estimate of heritage language abilities. However, task construction should consider morphophonological aspects vulnerable in heritage Arabic.