Morbidity in plague epidemics is an almost totally neglected field of research. In the present paper the evidence is presented and related to problems currently debated. The evidence clearly indicates that plague has a strong capacity for diffusion, leading to average morbidity rates in Tuscany 1630–2 (seven communities) and southern France 1720–2 (33 communities) of 44 and 55 per cent respectively. Combined with a lethality rate of close to eight per cent as in France, an epidemic pattern emerges which may serve to explain the reduction of the population during the late middle ages. Morbidity rates did not increase with increasing population densities, but were highest in small village communities, lowest in (small) towns and at an intermediate level in cities. This pattern has previously been observed in India, but no explanation was found. It is shown that the rat-based epidemiological model could serve to explain this peculiar pattern. This conclusion is supported by the few studies in which the spread of plague throughout local communities has been analysed.