This paper reviews the development of thinking about student dropout in general and also how this thinking has been adapted in the field of open and distance learning. It looks critically at the two basic research approaches: surveys of dropouts to ask them for their reasons, and studies that looked at progress in relation into likely predictive variables. Then it considers the seminal work of Tinto and examines two pieces of research in open and distance learning that have attempted to empirically test his longitudinal model. In conclusion, a case is made for interventionary tactics rather than a grand theory of such a multi-faceted phenomenon. [ABSTRACT FROM AUTHOR]