Purpose: The purpose of this paper is to argue that taking the educational purposes of schools into account is central to understanding the place and importance of facilities to learning outcomes. The paper begins by observing that the research literature connecting facility conditions to student outcomes is mixed. A closer examination of this literature suggests that when school facilities are measured from an engineering perspective, little connection to learning outcomes is evident. By contrast, when school facilities are rated in terms of educational functions, a connection to learning outcomes is apparent. Design/methodology/approach: The paper provides an empirical test of the educational relevance of how school facilities are measured. Using the schools in a Canadian division, the condition of school facilities was measured in two ways, including both conventional, engineering tools and a survey capturing principals' assessments. School facility ratings using these alternate measurement methods were correlated with schools' quality of teaching and learning environments (QTLE). Findings: Two central findings emerge. First, engineering assessments of facilities are unrelated to the QTLE in schools. Second, educators' assessments of school facilities are systematically related to the QTLE in schools. Originality/value: The findings indicate that more research needs to be directed at developing sound tools for measuring school facilities in terms of their educational relevance. In addition, school administrators need to reconsider policies that devalue the contribution that facilities make to learning outcomes. (Contains 4 tables and 6 notes.)