In many instances measures developed for the control of coffee berry disease in East Africa have become less effective since 1961. Changes in the flowering and fruiting patterns of the coffee tree have recently become evident, ascribed in part to the onset of a wetter and cooler climatic cycle in 1961–62, and in part to the effect of the disease itself which, by removing crop, stimulates out-of-season flowering. Consequently, instead of a single main crop in which susceptible and highly infective ripe berries are present for a short time only, a series of over-lapping crops results in ripe berries being on the trees for most of the year. This has changed the disease epidemiology: formerly the main source of infection was conidia from the maturing wood, but now conidia from the berries are often predominant. This has greatly affected the efficiency of control measures, and means are suggested for restoring the previous situation.