Although the Latin alphabet relies on phonological representation, almost all writing systems that are based on the Latin script include spelling with no representation in phonology. Writing, then, refers to morphological or grammatical structures that are not transparent in speech. Orthographic forms unrelated to speech are, indeed, significant cognitive challenges and, therefore, the most difficult to learn. Among them are plural markers in French and nominal capitalisation in German. The learners must learn these grammatical structures as they are related to writing in order to spell correctly. The overall objective of the GRASP project is to investigate the way explicit teaching of grammatical reflection en¬hances the spelling skills of Grade 5 pupils specifically in relation to orthographic markers that are not verbally expressed but are salient in writing. This objective is divided into three research questions, each related to a different perspective of the study. The questions examine the following areas: (1) short and long term training effects throughout Grade 5, (2) intra-individual spelling development, and (3) eventual transfer effects of the training in one language on spelling of the non-trained language. The project consists of two interrelated intervention studies, one for German and one for French, with parallel experimental designs that address these questions. The research project has great potential on a societal level. It addresses the fact that spelling mistakes, especially in grammatical spelling when the written markers cannot be observed in speech, have increased with recent generations of pupils. A strong grasp on grammatical structures is, however, a core factor for writing development, reading comprehension, and academic language skills. Spelling skills carry a certain importance socially as they form an integral part of academic achievement and affect pupils’ transition into secondary school. The context of the research project is linguistically and socially complex, with an officially bilingual school system, German and French, neither of these school languages refering to the family languages of the majority of the pupils. Beyond the Luxembourgish context, the project outputs will improve the understanding of spelling development of non verbalised but salient grammatical markers in writing among multilingual pupils.