Abstract. This paper examines reading as done by programmers engaged in software development. Reading is an activity we feel should be of fundamental interest to studies of programming, but the practical achievement of which has not been closely examined. We give examples of programmers reading in pairs, and reading alone, and show reading in both cases to be explainable in terms of shared social practices. These practices are not determined by the code but nor are they purely socially constructed; rather they lie in the linkage between the code and programmers ’ ways of reading the code. We discuss (1) how features of day-to-day coding work create pertinent occasions for reading a certain piece of code, (2) how programmers order and expect there to be an order to code, and (3) how programmers have ways of analysing code in order to make sense of it. This is an ethnomethodological study that draws from ethnographic fieldwork at a professional software development company.