Generic language about social categories risks amplifying essentialist beliefs, but its influence can be reduced by framing category features as products of extrinsic structures. We investigated which elements of essentialism are fostered by generics and whether structural framing dampens essentialist assumptions even for categories negatively stereotyped in society. In Experiment 1, participants read generic, quantified, or specific statements about novel categories and then rated the categories on ten dimensions of essentialism. Compared to other statements, generics led participants to essentialize the referenced categories broadly, viewing them as natural kinds with high inductive potential. In Experiment 2, participants read generics about novel or real-world categories (e.g., Mexican immigrants), sometimes accompanied by a structural explanation. Such explanations reduced essentialist interpretations of the generics for novel categories, but slightly increased them for real-world categories. Generics appear to induce broad essentialist beliefs, and structural framing may be insufficient for mitigating their problematic social consequences.