There has been a renewed interest in the philosophical and scientific basis of epidemiology in recent years. In particular, it has been argued that Popper's philosophy should be adopted by epidemiologists, an assertion that has met with some scepticism. However, most criticisms of Popper's approach have been from an inductivist viewpoint, concerned with the generation of theories, whereas Popper's concern is with the testing of theories, and the two schools have been largely "talking past each other". We present a critique of Popper from within his own domain of interest. Examples are presented to show that Popper's philosophy is incomplete even for the physical sciences on which it is based, and that it is particularly inappropriate for epidemiology. Popper's approach makes sense only under the narrow way he has chosen to define science, and thus provides only a possible answer to a small set of fundamental problems of science and its use in society. The recent Popperian "trend" has a positive aspect in that it has fostered deductive thinking, and exposed the shortcomings of induction. However, the restrictive Popperian framework actually inhibits discussion despite its veneer of "critical discussion". A more pluralistic approach is needed at this stage of the development of epidemiology.