In this study, further development of the Golden Rule (GR) approach is presented for the dynamics of intramolecular (IM) hydrogen atom and proton transfer (PT) in solution. The IM modes are treated following the procedure reported earlier which simplifies drastically the problem of evaluating the multidimensional transfer integrals. The polar solvent is treated as a dielectric continuum with classical Debye spectrum. In the most typical case of relation between the parameters involved, the rate constant is expressed as a product of two almost independent terms : the "pure" tunneling rate of the same transfer but without any reorganization effects taken into consideration, and a suppressing tunneling factor of Levich-Dogonadze type in a generalized form. Two major effects are present : the promoting effect of the IM vibrations symmetrically coupled to the reaction coordinate, and the suppressing effect resulting from the final reorganization of-both the molecule and solvent. This approach is applied to the hydrogen atom and proton transfer in the photochemical cycle of 5,8-dimethyl-1-tetralone (DMT) observed by Grellmann and co-workers in a polar protic solvent (EPA). This compound exhibits typical non-Arrhenius temperature and isotope dependence of the rate of triplet enolization. The kinetic curves of the ground-state reketonization reaction are close to Arrhenius, with significantly higher slopes than for typical intramolecular PT reactions. Semiempirical quantum-chemical calculations at AM1 level were carried out to study the relative stability, structure, and charge distribution of all states involved in the photochemical cycle, including the effects of solvation in a polar II-bonding solvent. Two rotamers E(I) and E(II) for the enol form were located corresponding to different positions of the H atom of the hydroxyl group. In ground state the first is more stable in both the gas phase and polar protic solvent modeled by water. Therefore, the reketonization reaction is treated as one-step tunneling from the rotamer E(I) to the keto form, i.e., without activated rotational equilibration E(I) <-> E(II) proposed by Grellmann and co-workers in an earlier study. Calculations of the rate constants were performed for both the direct and reverse reaction. Standard AM1 output (structural and force field data) was used as input, and good agreement with the available kinetic experiments was reached for both compounds. The high slope of the kinetic curve of this reaction is attributed to the additional activation energy resulting from the final reorganization of the low-frequency oscillators, mainly those from the solvation layer.