This study aims to provide information on the fate of triplet energy in DNA by considering the triplet-mediated reactions which occur between nucleic acid bases in solution, at room temperature. Following sensitization of base triplet states by acetone, the subsequent photophysics and photochemistry are highly dependent on the nature of the nucleotide pair under study. By establishing the direction of triplet energy transfer between pairs of mononucleotides, the relative triplet energy ordering was determined under physiologically relevant conditions, i.e. aqueous solution at room temperature. This order (C > U > G > A > T) is in agreement with that previously reported which was obtained at 77 K in rigid media. The absolute value for TMP triplet energy has been estimated as 310 kJ mol(-1), based on triplet energy transfer studies involving acetophenone and 3-methoxyacetophenone. Determination of the triplet energy gaps between all mononucleotides, to within 1 kJ mol(-1), has allowed an estimation of the absolute values of CMP (321 kJ mol(-1)), UMP (320 kJ mol(-1)), GMP (317 kJ mol(-1)), and AMP (314 kJ mol(-1)). The energy gaps between the triplet states are smaller than those reported at low temperature. This allows triplet energy transfer equilibria to be established in which a significant proportion of both triplets is present, unless thymine is present, in which case the dominant process is triplet energy transfer to the thymine. In any purine/pyrimidine (not thymine) pair, electron transfer from purine to pyrimidine is significant, producing the purine radical cation, which rapidly deprotonates to give the neutral radical. This process is most efficient for the guanosine-uracil combination and our evidence suggests that it is the purine triplet state which initiates the electron transfer. In the guanosine-adenosine system, there is no evidence of electron transfer and net triplet energy transfer is low due to a small triplet energy gap. This results in a relatively high proportion of both triplet states being observed in the equilibrium. These results show that DNA photochemistry should be highly sequence dependent and may have significance regarding the existence of "hot spots" in DNA photoproduct formation.