Prompted by discovering a paper written in 1980 which described a process of 'ethnography by proxy', I revisit this concept in the light of two research projects: a workplace study of an electronics company conducted in the early 1990s and a later home-based study of young children's encounters with toys and technology. The paper defines ethnography by proxy as the process of delegating some of the ethnographer's activities to participants in the research setting. It discusses a pragmatic response to some of the challenges of conducting fieldwork and considers the implications of delegating non-academic proxies to fulfil aspects of the ethnographer's role, the different guises that may be taken by the ethnographer and what this shift in research relationships might mean for the interpretation of data. The concept of proxy has implications for some of the tenets of ethnographic research as it requires a re-examination of our roles and the relationship between researcher and researched.