Carbon-based nanostructures have attracted tremendous interest because of their versatile and tunable properties, which depend on the bonding type of the constituting carbon atoms. Graphene, as the most prominent representative of the π-conjugated carbon-based materials, consists entirely of sp2-hybridized carbon atoms and exhibits a zero band gap. Recently, countless efforts were made to open and tune the band gap of graphene for its applications in semiconductor devices. One promising method is periodic perforation, resulting in a graphene nanomesh (GNM), which opens the band gap while maintaining the exceptional transport properties. However, the typically employed lithographic approach for graphene perforation is difficult to control at the atomic level. The complementary bottom-up method using surface-assisted carbon–carbon (C–C) covalent coupling between organic molecules has opened up new possibilities for atomically precise fabrication of conjugated nanostructures like GNM and graphene nanoribbons (GNR), although with limited maturity. A general drawback of the bottom-up approach is that the desired structure usually does not represent the global thermodynamic minimum. It is therefore impossible to improve the long-range order by postannealing, because once the C–C bond formation becomes reversible, graphene as the thermodynamically most stable structure will be formed. This means that only carefully chosen precursors and reaction conditions can lead to the desired (non-graphene) material.