A wide range of experimental data from earlier studies by other workers are combined with recent data from the Burrows group to interpret that group's thymine dimer (T = T) repair rate data for 8-oxo-7,8-dihydroguanine (OG)-containing DNA duplexes. The focus of this effort is to explain (i) how and why the repair rates vary as the sequence location and distance of the OG relative to the T=T is changed and (ii) why the spatial extent over which repair is observed is limited to OG-T=T distances of similar to 6 angstrom. it is proposed that, if the OG and T=T are within similar to 5-6 angstrom, a Coulomb potential moves the energy of the OG(+)center dot center dot center dot T=T- ion-pair state below the photoexcited OG*center dot center dot center dot T=T state, even in the absence of full solvent relaxation, thus enhancing forward electron transfer from OG* to T=T by allowing it to occur as a radiationless internal conversion process rather than by overcoming a solvation-related barrier. The rate of this forward electron transfer is estimated to be similar to 10% of the decay rate of the photoexcited OG*. For OG-to-T=T distances beyond 5-6 angstrom, electron transfer is still exothermic, but it must occur through solvent reorganization, overcoming an energy barrier, which presumably renders this rate too slow to be detected in the experiments under study here. Once an electron has been injected into the T=T, as many other workers have shown, the reaction proceeds through two low-energy barriers first connecting T=T- to an intermediate in which the C-5-C-5, bond of the cyclobutane unit is cleaved, and onward to where the cyclobutane unit is fully broken and two intact thymine sites are established. Our ab initio data show that the energy landscape for these bond cleavages is altered very little by the presence of the proximal OG(+) cation, which therefore allows us to use data from the earlier studies to conclude that it takes similar to 100 ps for complete bond cleavage to occur. The experimentally determined overall T=T repair quantum yield of 1% then allows us to estimate the rate at which an electron is transferred from the T=T- anion back to the OG+ cation as 10 times the rate of bond cleavage. The experimental variations in T=T repair rates among several sequences are shown to be reasonably consistent with an exponential OG-to-T=T distance dependence, e(-beta R), with a decay parameter of beta = 0.6 angstrom(-1). Finally, suggestions are offered for experimental studies that would test the predictions offered here and shed further light on the OG-induced T=T repair mechanism.