Patients throughout Europe are concomitantly exposed to multiple pollens from distinct Pooideae species. Given the overlap in pollination calendars and similar grain morphology, it is not possible to identify which grass species are present in the environment from pollen counts. Furthermore, neither serum IgE reactivity nor skin prick testing allow the identification of which grass species are involved in patient sensitisation. Due to their high level of amino acid sequence homology (e.g., >90% for group 1, 55–80% for group 5), significant cross-immunogenicity is observed between allergens from Pooideae pollens. Nevertheless, pollen allergens also contain species-specific T or B cell epitopes, and substantial quantitative differences exist in allergen (e.g., groups 1 and 5) composition between pollens from distinct grass species. In this context, a mixture of pollens from common and well-characterised Pooideae such as <i>Anthoxanthum odoratum, Dactylis glomerata, Lolium perenne, Phleum pratense </i>and <i>Poa pratensis </i>is suitable for immunotherapy purposes because (1) it has been validated, both in terms of safety and efficacy, by established clinical practice; (2) it reflects natural exposure and sensitisation conditions; (3) it ensures a consistent and well-balanced composition of critical allergens, thus extending the repertoire of T and B cell epitopes present in the vaccine.