Much pedagogic research undertaken in geography and other disciplines relies on post-hoc methods such as surveys or interviews to investigate the student experience of higher education (often based on self-reports of behaviour). However, observation of students provides a far more direct route to obtain information about their behaviour, and there are many situations where observational data--collected by a researcher or by students themselves using video, audio or written diaries--may provide a deeper insight into their experience. This paper outlines some basic methods of observation, describes the potential advantages and disadvantages of collecting observational data, and provides examples of relevant research which draws on these methods. We conclude with a recommendation that observational methods should be used more widely in pedagogic research.