In this work, we present DFT calculations of the energetics of the base-pair anion and cation radicals of adenine-thymine (AT) and adenine-uracil (AU). At the B3LYP/6-31+G(d) level, we find that the adiabatic electron affinities (AEAs) are 0.30 eV for AT and 0.32 eV for AU. These values are both positive but slightly smaller than previously reported values for the AEA of guanine-cytosine (GC) and the hypoxanthine-cytosine base pair (IC). Furthermore, the AT and AU anion radical vertical electron detachment energies are also smaller than those of GC and IC, with that of AT only about half of GC's. For electron transfer between two identical isolated base pairs,. the reorganization energies, lambda, are calculated to be AT(0.76), AU(1.06), IC(1.25), and GC(1.31 eV). These results indicate that the AT base pair has a shallow trap depth and provides a favorable route for electron transfer, which explains previous experimental results that electron-transfer rates were higher in polydAdT than in polydGdC. Values of the ionization energies reported are in good agreement with the best estimates of previous work. For hole transfer, the reorganization energies, lambda, are calculated to be AT(0.37), AU(0.53), IC(0.66), and GC(0.70 eV). These suggest that hole transfer through sequences of stacked AT base pairs may be most favorable. Base-pairing energies are also reported, which show that the formation of cation and anion radicals tends to increase base-pairing energies substantially, with cations more strongly affected than anions. We further show that whereas the predicted electron affinity of the individual base hypoxanthine (abreviated "I") is very slightly more than that of cytosine, base pairing in IC increases the relative electron affinity of cytosine in relation to that of hypoxanthine. Thus, we find that as I approaches C the electron transfers from I to C so that the electron localizes preferentially on cytosine in the fully optimized IC base-pair anion radical.