Background: Understanding and expressing a narrative's macro-structure is relatively independent of experience in a specific language. A narrative task is therefore assumed to be a less biased method of language assessment for bilingual children than many other norm-referenced tests and may thus be particularly valuable to identify language impairment (LI) in a bilingual context. Aims: The present study aimed to investigate the effects of LI and bilingualism on macrostructural narrative skills. Moreover, it evaluated the diagnostic validity of a narrative task within a monolingual and bilingual sample. Methods & Procedures: Five- and 6-year-old mono- and bilingual children with and without LI (N = 33 per group) were tested on production and comprehension measures of narrative macro-structure. A multivariate analysis of covariance (MANCOVA) was used to investigate the effects of LI and bilingualism on their narrative abilities. Binary logistic regressions were conducted to evaluate the instrument's diagnostic value. Outcomes & Results: Negative effects of LI were found on all narrative measures, whereas no effects of bilingualism emerged. The narrative task adequately differentiated between both mono- and bilingual children with and without LI, with story elements related to internal states being more effective than elements related to the basic episode structure. Conclusions & Implications: This study confirms the hypothesis that measures of narrative macro-structure are not biased against children who have less experience with a particular language, like bilinguals. In addition, it indicates that using narratives to assess children's language abilities can support the identification of LI in both a monolingual and a bilingual context.