Much of the research evidence relating to the physical learning environment of schools is inconclusive, contradictory or incomplete. Nevertheless, within this confusing area, research from a number of disciplines, using a range of methodologies, points to the negative impact of noise on students’ learning. In this paper, drawing on our systematic review of learning environments we review the weight of evidence in relation to noise, considering what implications the results of these studies have for the design and use of learning spaces in schools. We make four key points. Firstly that noise over a given level does appear to have a negative impact on learning. Secondly that beneath these levels noise may or may not be problematic, depending on the social, cultural and pedagogical expectations of the students and teachers. Thirdly we argue that when noise is deemed to be a difficulty, this finding cannot simply be translated into design prescriptions. The reasons for this indeterminacy include differing understandings of the routes through which noise produces learning deficits, as well as relationships between noise and other elements of the environment, particularly the impacts of physical solutions to noise problems. Finally, we suggest that solutions to noise problems will not be produced by viewing noise in isolation, or even as part of the physical environment, but through participatory approaches to understanding and adapting the structure, organisation and use of learning spaces in schools.