In demographic studies on Brassica cultivars and accessions we observed large genotypic variation in the ability of seedlings to establish. Here we quantify the role of slugs during establishment by including and excluding slug herbivory.Slug exclusion during the first 10 days after germination led to 26.5 times more surviving seedlings in modern canola varieties of Brassica napus, 4.3 times more in feral B. napus and 1.9 times more in wild Brassica rapa. While seed and seedlings of canola were low in aliphatic glucosinolates (AGS), feral B. napus and wild B. rapa had much higher AGS levels. Consequently, we find a positive correlation between the survival of young seedlings and AGS concentration. Concentrations of indole glucosinolates (IGS) in seeds were much lower than those of AGS and did not correlate significantly with survival. Subsequently, we exposed 10-day-old seedlings to herbivory. In the 4 days following exposure, slug herbivory was negatively correlated to AGS concentration but not to IGS.In choice experiments in the lab, the preference of the slug Arion lusitanicus also correlated negatively with AGS content. A. lusitanicus preferred to feed on B. napus rather than on wild B. rapa, but had no significant preference when presented with plants similar in AGS content.Slugs can be a limiting factor for seedling recruitment in populations of B. napus, especially for modern canola cultivars with a low AGS content.