Coronene is one of the basic polycyclic aromatic hydrocarbons (PAHs) used to test the reliabilty of a multidimensional potential energy surface (PES) and to assess its influence on the formation dynamics of PAH clusters with defined physical and chemical properties. We report an analytical potential energy surface for modeling the coronene–coronene interaction, whose parameters were fine-tuned on dispersion-corrected DFT estimations performed within the generalized gradient PBE approximation, that is suitable for describing molecular aggregates involving aromatic species. This model was used to build a potential function for coronene clusters (Cor n ) that was then applied in a detailed global geometry optimization study with an evolutionary algorithm. A large variety of low-energy structures were obtained for the Cor n (n = 2–15) clusters ranging from columnar-type to two-stacked in a handshake association motifs. Moreover, it was found that a transition from a single-stack columnar regime to other more complex shapes occurs at n = 6, whereas previous results based on a simpler coarse-grained potential pointed to a transition at n = 8. Geometry reoptimizations were also performed at the DFT level for the most representative low-energy structures of Cor n (n = 3–6), which confirmed the reliability of the present findings.