To identify energy traps in CP43, a subcomplex of the photosystem II antenna system, site energies and excitonic couplings of the QY transitions of chlorophyll (Chl) a pigments bound to CP43 are computed using electrostatic models of pigment–protein and pigment–pigment interactions. The computations are based on recent crystal structures of the photosystem II core complex with resolutions of 1.9 and 2.1Å and compared to earlier results obtained at 2.9Å resolution. Linear optical spectra (i.e., absorption, linear dichroism, circular dichroism, and fluorescence) are simulated using the computed excitonic couplings, a refinement fit for the site energies, and a dynamical theory of optical lineshapes. A comparison of the obtained root mean square deviation of about 100cm−1 between directly calculated and refined site energies with the maximum range of about 350cm−1 of directly calculated site energies shows that the combined quantum chemical/electrostatic approach provides a semi-quantitative agreement with experiment. Possible reasons for the deviations are discussed, including limits of the electrostatic models and the lineshape theory as well as structural alterations of CP43 upon detachment from the core complex. Based on the simulations, an assignment of the two low-energy exciton states A and B of CP43, that where observed earlier in hole burning studies, is suggested. State A is assigned to a localized exciton state on Chl 37 in the lumenal layer of pigments. State B is assigned to an exciton state that is delocalized over several pigments in the cytoplasmic layer. The delocalization explains the smaller inhomogeneous width of state B compared to state A observed in hole burning spectra, which is proposed to be due to exchange narrowing. The assignment of states A and B largely confirms our earlier suggestion that was based on a fit of linear optical spectra and electrostatic calculations using the 2.9Å resolution structure. Interestingly, for the latter structure, the site energy of Chl 37 is obtained closer to the refined value than for 1.9 and 2.1Å resolution. This is explained by a variation of the site energy due to the influence of lipids that might be different in the core complex and isolated CP43. To remove remaining uncertainties in the assignment of states A and B, target sites for mutagenesis experiments are proposed based on the electrostatic computations. In particular, it is suggested to mutate Trp C63 close to Chl 37 to probe the identity of state A and to mutate Arg C41 close to Chl 47 to probe state B.