Ab initio MODPOT/VRDDO/MERGE SCF calculations have been carried out for the attack of CH3+ on O6 and N7 of guanine (for both planar and pyramidal CH3+). There are minima in the potential energy surfaces at about 3 bohr. As expected, there is a crossing of the potential energy surfaces for planar and pyramidal CH3+ attack at ca. 3.5–3.75 bohr, since at the dissociation asymptote CH3+ planar is more stable, but attached to guanine CH3+ pyramidal would be more stable. The most interesting and fundamental finding from our calculations was not anticipated in advance. CH3+ + guanine is an ion-molecule reaction. In ion-molecule reactions if the reactants A+ + B are higher in energy at the asymptote than A + B+, then there will be a crossing or avoided crossing of the potential energy surfaces of A+ + B and of A + B+. In our SCF calculations of CH3+ + guanine at about 5.0 bohr there began a mixing of electronic configurations. This indicated that the ionization potential of guanine (which had not previously been reported as measured experimentally) had to be lower than that of CH3+, 9.8 eV. Unpublished photoelectron spectra of Le Breton indicated that the ionization potential of guanine was ca. 8.3 eV. The ionization potential of the methyl radical is higher than that of all the DNA bases, and thus there will be mixing of potential energy surfaces. Only ab initio configuration interaction (or ab initio MODPOT/VRDDO/CI or ab initio MODPOT/VRDDO/MERGE/CI) calculations can properly describe such a reaction pathway. On the contrary, the ionization potential of the ethyl radical, 8.4 eV, is lower than that of thymine and cytosine, and about that of adenine and guanine. Thus, completely different sets of potential energy surfaces will be available to the Et+ + DNA base reactions. The first two Et+ + DNA base reactions might be able to proceed along a lowest-energy singlet single-determinant surface. The latter two will have strong configuration mixing. Again, only CI calculations can resolve these reaction pathways.