The delayed fluorescence properties of proflavin have been exploited in studies of the excited-state binding kinetics of the dye to poly[d(A-T)] and its brominated analogue poly[d(A-br5U)] at room temperature and pH 7. The two analyzed luminescence decay times of the DNA-dye complex are dependent on the total nucleic acid concentration. This dependence is shown to reflect a temporal coupling of the intrinsic delayed emission decay rates with the dynamic chemical kinetic binding processes in the excited state. Temperature-jump kinetic studies conducted on the brominated polymer and corresponding information on poly[d(A-T)] from a previous study [Ramstein, J., Ehrenberg, M., & Rigler, R. (1980) Biochemistry 19, 3938-3948] provide complementary information about the ground state. In the ground state, the poly[d(A-T)]-proflavin complex has one chemical relaxation time, which reaches a plateau at high DNA concentrations. The brominated DNA-dye complex exhibits two relaxation times: a faster relaxation mode that behaves similarly to that for the unhalogenated DNA and a slower relaxation mode that is apparent at high DNA concentrations. The ground-state kinetic data are analyzed in terms of two alternative models incorporating series and parallel reaction schemes. The former consists of two sequential binding steps--a fast bimolecular process followed by a monomolecular step--while the latter consists of two coupled bimolecular steps. A similar analysis for the excited-state data yields reasonable kinetic constants only for the series model, which, in accordance with previous proposals for acridine intercalators, consists of a fast outside binding step followed by intercalation of the dye. A comparison of the ground- and excited-state kinetic parameters reveals that the external binding process is much stronger and the intercalation is much weaker in the excited state. That the excited-state data are only consistent with the series model suggests that delayed luminescence studies may provide a general tool for distinguishing between the two kinetic mechanisms. In particular, we demonstrate the use of delayed luminescence spectroscopy as a tool for probing dynamic DNA-ligand interactions in solution.