(1) A 2-year demographic study was made of four adjacent populations of Hieracium pilosella on contrasting soil types in Breckland, in order to examine whether the population fluxes were different. Three of the four sites were rabbit-grazed, but at the fourth a rabbit-proof exclosure was constructed. (2) Recruitment of rosettes to all populations was the result of clonal growth, as seedling establishment failed to occur. However, sexual reproduction and clonal growth were closely linked; it was observed that an immature inflorescence needed to be produced before the production of stolons and the development of rosettes at the ends of the stolons could occur. Plants senesced after undergoing reproduction. (3) All the populations showed overall exponential decay rates, upon which was super-imposed an annual rhythm of natality and mortality. Initially all four populations showed fairly similar overall exponential decay rates, but midway through the study the rate of decline abruptly increased in the populations growing in acidic sand and chalk soil, with the result that one population became almost extinct, and in the other the age of the majority of individuals in the population consisted of young plants less than 0.5 years of age. (4) The absence of grazing resulted in the population in the exclosure consisting of a greater proportion of plants 2 years or older compared with the other sites. (5) The significance of these differences in the population fluxes at the sites is discussed in relation to changes in Breckland populations of H. pilosella reported earlier by Watt (1962), and to the exceptional drought of 1975-6.