Abstract There has been increased focus on the explicit teaching of grammatical knowledge since the new English National Curriculum (2014) and the mandatory grammar test; but little is known about how grammar is learnt. This study set out to investigate children’s and teachers’ metalinguistic understanding of closely-related grammatical concepts: verb, tense, aspect, modality and voice in English grammar. It also considered how children used these grammatical patterns in writing. This appears to be the first such research in the English-speaking world. Metalinguistic Achievement Tests and interviews were designed for Y2 and Y5/6 children (n=139); also, teachers (n=10) completed a grammar test and were interviewed to establish the nature of their understanding. A number of questions were investigated, including: To what extent can children and teachers understand the choices of tense, aspect, modality and voice? The findings indicated some similarities between children’s and teachers’ understanding. Verbs were defined and understood as 'doing words'; and so excluded state verbs. The perfect and progressive aspects were confused with the past and present tenses implied by the past and present participles. Their understanding of passives compared to active was conceptualised by a ‘Swap it around/Going the other way around’ notion, but without reference to their syntax. In writing, although the children used complex patterns such as aspect and passive, there was limited metalinguistic knowledge and/or understanding. The thesis offers a conceptual distinction between ‘knowledge’ and ‘understanding’: some children/teachers could identify linguistic form, but did not have the grammatical understanding. Two conceptual terms are introduced, grammatical metalinguistic knowledge and grammatical metalinguistic understanding. Various recommendations on future research, practice and policy are made, including a longitudinal study, tracking children’s progress from Y2 to Y6; and reviewing the assessment of grammar, particularly on the efficacy of solely using ‘pencil and paper’ tests. It may help to develop strategies and professional development opportunities for in-service teachers – it is unlikely to be wise to expect teachers to learn the intricate web of grammar on their own.