Doctoral students have often been described as apprentices engaged in workplace learning. Further, assumptions are frequently made in the literature about the common nature of such learning experiences, e.g., in the sciences, research-related practices are learned in a lab within the supervisor’s program and team. A few recent studies of the science doctoral experience have challenged this view arguing such assumptions may overlook considerable variation. This longitudinal study, using frequently completed activity logs and an interview, reports on the research-related practices of twelve UK science doctoral students. The analysis, particularly of the logs, challenged some of the literature-based assumptions: students often chose to work in institutional offices, non-institutional sites and their homes rather than in labs; they did not necessarily engage regularly with a research team, nor were they necessarily engaged in a project directly linked to their supervisors’. That students chose not to work in traditionally assumed places suggests the importance of attending to: a) student agency, b) how research-related practices may be changing, and c) how sites of doctoral learning might need to be reconceived. As well, the findings suggest the value of non-traditional data collection methods in capturing variation in experience.