The aim of this paper is to defend the structural concept of representation, as defined by homomorphisms, against its main objections, namely: logical objections, the objection from misrepresentation, the objection from failing necessity, and the copy theory objection. The logical objections can be met by reserving the relation ¿to be homomorphic to¿ for the explication of potential representation (or, of the representational content). Actual reference objects (¿targets¿) of representations are determined by (intentional or causal) representational mechanisms. Appealing to the independence of the dimensions of ¿content¿ and ¿target¿ also helps to see how the structural concept can cope with misrepresentation. Finally, I argue that homomorphic representations are not necessarily ¿copies¿ of their representanda, and thus can convey scientific insight.