SUMMARYFour experiments were done under glass in which tomato plants were inoculated with TMV (tomato strains of tobacco mosaic virus) at different times during their growth, and most of the healthy control plants were kept uninfected throughout the season. Leaf symptoms were visible 2–3 weeks after inoculation, and infected plants were checked and grew more slowly than healthy ones. Leaves of infected plants were sometimes scorched by overhead damping with water, and lower leaves died earlier than those of healthy plants.In the 1960 experiment, with a winter‐sown crop, plants inoculated on 3 March, when the first trusses began to flower, suffered a 15% loss in weight of fruit; those inoculated on 14 April, when the fifth trusses began to flower, an 18% loss, and those inoculated on 31 May, when most of the fruits had set, a 14% loss. The two early infections caused 13% fewer fruits to set, but the last one only 3% fewer; however, with successively later infection dates, mean fruit size progressively decreased. Fruit quality was severely affected by later infection, 10% of the fruits from plants infected in April showing necrotic pitting, bronzing or severe mottle, but 31% of those infected on 31 May.Three experiments with spring‐sown plants gave similar results. In the 1961 experiment, plants inoculated on 3 May, when the first trusses began to flower, yielded 21% less weight of fruit than the controls, and those inoculated on 8 June, when the fifth trusses began to flower, 14% less. A tobacco strain of TMV that caused no symptoms in tomato failed to protect plants from later infection with a tomato strain. In the 1962 experiment, seedlings inoculated on 13 March, just after pricking out, yielded 10% less weight of fruit; plants inoculated on 4 April, just after planting out, 3% less; those inoculated on 3 May, when the first trusses began to flower, 11% less; and those inoculated on 7 June, when the fifth trusses began to flower, 16% less. In the 1963 experiment, plants inoculated on 21 May, when the first trusses began to flower, yielded 17% less weight of fruit than the controls, and those inoculated on 20 June, when the fifth trusses began to flower, yielded 23% less; many of the control plants became infected during August in this last experiment. As in the first experiment, losses were largely due to poor setting of fruits at the time the plants became systemically infected with virus, and necrotic pitting and severe bronzing again mainly affected fruits from the late‐infected plants. The plants grew vigorously and were ‘soft’ or lush in all the four experiments.