Health literacy encompasses people’s knowledge, motivation and competences to access, understand, appraise and apply health information to make judgments and take decisions in every-day life concerning health care, disease prevention and health promotion (Kickbusch et al., 2013). Research in health literacy has grown in prominence as a separate entity over the last twenty years, partly in response to the growth of chronic disease requiring greater patient self-management. Research has found significant relationships between reading comprehension levels and different health outcomes and health behaviours, with reading comprehension being used as a way to measure health literacy despite its accepted definition being much broader. However, there has been little research which takes a social practices perspective on health literacy. In this thesis, health literacy is viewed from a social practices perspective. Literacy events in this research have been predicated on the existence of written texts as part of a patient’s journey using health related texts, resources and services. The thesis explores how people go about accessing, understanding, appraising and using health-related services and information, and the social and cultural practices inherent in those activities. It asks how much written patient text is used with patients in hospital; who is using it; what the discourse features of those texts are; and how patients respond to the texts. This mode of exploration is called textual ethnography (Swales, 1998). The research takes an ethnographic perspective not only on text, but also on participants and their journey. It shadows eleven patients through a cardiology service of a hospital in New Zealand. It identifies literacy events and practices occurring in that space and examines the patient texts themselves using a discourse analysis (Bax 2011; Faircough, 2003; Clerehan, Hirsh, & Buchbinder, 2009; Franken & Hunter, 2011) and geosemiotic (Scollon & Scollon, 2003) approach. The key themes to emerge from the research centre around the complexity of access to patient information, both physically and linguistically, and around the dynamics of patient agency. The ability to encode and decode written text played a minor part in patient access to information. The analysis showed that the amount and types of text material present for patients in hospital were limited. Access to key texts required a deliberate decision by a health professional to dispense the text items. Mediation of text in the research by literacy sponsors supported Brandt and Clinton’s (2002) observation that “access to literacy has always required assistance, permission, sanction, or coercion by more powerful others or, at least, contact with existing ‘grooves’ of communication” (p. 349). Patient agency therefore came to the fore, and practices that patients did actually use to make information personal to them such as talking with room mates, or requesting paper and pencil were dynamic and socially mediated. Implications lie in the way health literacy is conceptualised, moving the focus from individual abilities as portrayed in current health literacy definitions, and pointing to the wider role of the health organisation and health care practitioners as literacy sponsors. Practice implications arising from this reconceptualisation relate to patient information strategies within health care organisations and pedagogical implications for health professionals and adult literacy educators. The notions of text mediation and full contextualisation of learning also became apparent with evidence pointing to the effectiveness of text mediation by a more knowing other. The hospital itself was found to be a potentially rich and meaningful setting for adult learning about health where full contextualisation of adult learning is possible. These patient learning opportunities need to be further exploited by health organisations and professionals.