‘In the last ten years on my individual reckoning’, observes S. E. Finer, ‘there have been seventy-three coups in forty-six countries’. ‘Coups’, Gurr comments, ‘can alter political processes and social institutions as drastically as any classic revolutions’. Yet the incidence and importance of coups are hardly reflected in the sparse literature proposing generalizations about their causes. Certainly, many case studies of individual coups have been undertaken, but the choice of the country has usually been decided by availability of data rather than its significance for a general theory. Given that coups have occurred all over the world, they clearly are a general phenomenon. Existing general explanations for them, however, are open to criticism. These suggest that essentially there is room for a theory which is about coups in particular rather than about wider forms of political instability, or about the narrower, military, coup; which is capable of falsification, avoiding inherently untestable hypotheses or concepts that are defined so loosely as to invite accusations of tautology; and which is, of course, able to withstand appropriate empirical examination.