(1) Two host-plants of Antestia lineaticollis in addition to Coffea arabica are Psychotria nairobiensis and Pavetta elliottii.(2) Laboratory experiments showed that red berries and large green berries are favoured foods, but feeding will also take place on small green berries, on shoots, and on leaves.(3) Feeding trials showed that large green berry food is essential for normal length of life and normal egg production.(4) Antestia was reared to adult stage when fed exclusively on large green berries, on small green berries, and on shoots, but not on red berries or leaves.(5) In large green berries and small green berries the endosperm becomes infected with rots, usually caused by the fungi Nematospora coryli or Nematospora gossypii, which are introduced by Antestia, but in the case of ripe red berries the endocarp is seldom if ever pierced and rotting of the endosperm is not initiated at this stage.(6) The amount of damage caused to the crop by rotting of the beans is related to the number of Antestia present, and a formula which is considered to represent this relationship is given. In practice as high a proportion as 97 per cent. of the beans have been found rotted when Antestia is numerous. The figures also show that even when Antestia population is low, an economically important amount of damage to the crop may occur.(7) Antestia was proved to cause a dropping of young green berries. This seems not to be due to the insect feeding on the stalk, but to be a response by the plant to an interruption of normal growth of the bean due to insect feeding. A similar shedding has been noticed of berries damaged by other insects when it is known that no feeding on the stalk can have taken place.(8) Feeding experiments in cages and sleeves showed that Antestia must rank as one of the important causes of the non-setting of flowers ; this form of damage has been neglected or minimised in the past.(9) As a result of Antestia feeding on the growing point, leaves become scarred and distorted.(10) A useful form of cage for experimental work of this kind is described.(11) Definite confirmation was obtained that well-known abnormalities of the vegetative growth characterised by multiple branching are produced by the feeding of Antestia.(12) It is shown that the growth is stopped or much reduced by the feeding of Antestia, and when this is associated, as it usually is, with multiple branching, a “ bunchy ” or “ matted ” growth results, which increases pruning costs and much reduces the cropping ability of the tree.(13) A short discussion follows which indicates certain practical conclusions that can be drawn from the above results.