Surprisingly little research has been published on the responses to elevated [CO2] at the community level, where herbivores can select their preferred food. We investigated the combined effects of atmospheric [CO2] and herbivory on synthesised plant communities growing on soils of different fertility. Factorial combinations of two [CO2] (350 or 700 [Formula: see text]l l-1), two fertility (fertilised or non-fertilised), and two herbivory (herbivores present or absent) treatments were applied to a standard mixture of seven fast- and eight slow-growing plants in outdoor microcosms. The herbivores used were the grain aphid (Sitobion avenae) and the garden snail (Helix aspersa). We measured plant biomass, foliar nitrogen and soluble tannin concentration, aphid fecundity, and snail growth, fecundity, and feeding preferences over one growing season. Elevated [CO2] did not have a significant impact on (1) the combined biomass of fast-growing or slow-growing plants, (2) herbivore feeding preferences, or (3) herbivore fitness. There was, however, a significant biomass increase of Carex flacca (which represented in all cases less than 5% of total live biomass), and some chemical changes in unpalatable plants under elevated [CO2]. The herbivory treatment significantly increased the biomass of slow-growing plants over fast-growing plants, whereas fertilisation significantly increased the abundance of fast-growing plants over slow-growing plants. Predictions on the effects of elevated [CO2] based on published single-species experiments were not supported by the results of this microcosm study.