This paper tests the prediction that introduced plants may become successful invaders because they experience evolutionary changes in growth and defence in their new range [evolution of increased competitive ability hypothesis (EICA)]. Interspecific and intraspecific binary feeding choices were offered to the snail Helix aspersa. The choices were between: (1) plants of the invasive Senecio inaequidens and Senecio pterophorus derived from populations in the introduced range (Europe) and plants of three indigenous species (Senecio jacobea, Senecio vulgaris and Senecio malacitanus) from populations in Europe; (2) plants of the invasive S. inaequidens and S. pterophorus from populations in the introduced range (Europe) and from populations in the native range (South Africa). We did not find a clear pattern of preference for indigenous or alien species of Senecio. However, we found that European invasive populations of S. inaequidens and S. pterophorus were less palatable than South African native populations. Moreover, in contrast to the predictions of the EICA hypothesis, the invasive genotypes of both species also showed a higher total concentration of pyrrolizidine alkaloids, and in the case of S. inaequidens we also found higher growth than in native genotypes. Our results are discussed with respect to the refinement of the EICA hypothesis that takes into account the difference between specialist and generalist herbivores and between qualitative and quantitative defences. We conclude that invasive populations of S. inaequidens and S. pterophorus are less palatable than native populations, suggesting that genetic differentiation associated with founding may occur and contribute to the plants' invasion success by selecting the best-defended genotypes in the introduced range.