In recent years, model-driven approaches and processes have1 established themselves as pragmatic and feasible solutions with tangible2 advantages. Transformations play a central role in any model-driven so-3 lution and, as interest in textual modelling grows, providing concepts and4 tools for supporting a high-level and declarative specification of bidirec-5 tional model-to-text transformations becomes a vital area of research.6 Our paper identifies important areas and scenarios for model-to-text7 transformations that are not or only partially supported by currently8 existing solutions. Based on the requirements of a real-world case study,9 we introduce a new concept that has been inspired by a successful bidi-10 rectional model-to-model transformation approach: Triple Graph Gram-11 mars.12 Key words: textual modelling, model-to-text transformations, bidirec-13 tional transformations, pair grammars, triple graph grammars, AST14