Invasion success of introduced species is often attributed to a lack of natural enemies as stated by the enemy release hypothesis (ERH). The ERH intuitively makes sense for specialized enemies, but it is less evident why invaders in their new area escape attacks by generalist enemies. A recent hypothesis explains low herbivore damage on invasive plants with plant defense chemicals that are evolutionarily novel to native herbivores. Support for this novel weapon hypothesis (NWH) is so far based on circumstantial evidence. To corroborate the NWH, there is a need for direct evidence through explicit characterizations of the novel chemicals and their effects on native consumers. This study evaluated the NWH using the highly invasive red alga Bonnemaisonia hamifera. In pairwise feeding experiments, preferences between B. hamifera and native competitors were assessed for four common generalist herbivores in the invaded area. Through a bioassay-guided fractionation, we identified the deterrent compound and verified its effect in an experiment with the synthesized compound at natural concentrations. The results showed that native herbivores strongly preferred native algae to B. hamifera. The resistance against herbivores could be tracked down to the algal metabolite 1,1,3,3-tetrabromo-2- heptanone, a compound not known from native algae in the invaded area. The importance of the chemical defense was further underlined by the feeding preference of herbivores for individuals with a depleted content of 1,1,3,3-tetrabromo-2- heptanone. This study thus provides the first conclusive example of a highly successful invader where low consumption in the new range can be directly attributed to a specific chemical defense against evolutionarily naïve native generalists. In conclusion, our results support the notion that novel chemical weapons against naïve herbivores can provide a mechanistic explanation for plant invasion success. © 2012 by the Ecological Society of America.