This paper begins by problematizing the `taken for granted' status of Newtonian linear time at the heart of (social) psychology. Borrowing from Adam, the paper makes a distinction between an `events in time' and `time in events' approach to social psychology. It argues that a `time in events' approach helps to reveal the importance of multiple times for social psychological theories and practices. To demonstrate this approach, it considers some of the multiple times that are relevant for analysing the concept of `identity'. It also explores the multiple times of the research encounter. In doing so, it suggests that traditional dichotomies in social psychology between synchronic and diachronic methods, and between experimental and qualitative methods, can be dissolved by this focus on the temporal. The paper concludes with some observations about the importance of time for thinking about social psychology as a knowledge-producing practice.