The success of invasive plants has been attributed to their escape from natural enemies and subsequent evolutionary change in allocation from defence to growth and reproduction. In common garden experiments with Senecio jacobaea, a noxious invasive weed almost worldwide, the invasive populations from North America, Australia, and New Zealand did indeed allocate more resources to vegetative and reproductive biomass. However, invasive plants did not show a complete change in allocation from defence to growth and reproduction. Protection against generalist herbivores increased in invasive populations and pyrrolizidine alkaloids, their main anti-herbivore compounds, did not decline in invasive populations but were higher overall compared with native populations. In contrast, invasive plants lost additional protection against specialist herbivores adapted to pyrrolizidine alkaloids. Hence, the absence of specialist herbivores in invasive populations resulted in the evolution of lower protection against specialists and increased growth and reproduction, but also allowed a shift towards higher protection against generalist herbivores.