There is confusion in the ecological literature between seed dormancy and persistence in soil. Some ecologists seem to assume that dormancy is necessary for persistence, while others imply that dormancy and persistence are virtually synonymous. Here, we show that there is no close relationship between dormancy and persistence and, incidentally, that conventional methods of investigating soil seed banks underestimate the persistence of species with dormant seeds. The confusion appears to arise from the concept of ‘enforced dormancy’, which is not genuinely dormancy at all, and would be eliminated if ecologists adopted the definition of dormancy employed by physiologists. Dormancy is a characteristic of the seed, not of the environment, the degree of which defines the conditions required to make the seed germinate.