The early suggestion by Lozier and Butler (Photochem. Photobiol. 17, 133-137 (1973)) that EPR Signal II arises from radicals associated with the water-splitting process in PSII has been confirmed and extended over the intervening years. Recent work has identified the Signal II radicals, [Formula: see text] and [Formula: see text], with plastosemiquinone cation species. In the experiments presented here we have used ENDOR spectroscopy and D2O/H2O exchange to characterize these paramagnets in more detail. The ENDOR matrix region, which arises from protons which interact weakly with the unpaired electron spin, is well-resolved at 4 K and at least seven resonances are apparent. A number of hyperfine couplings in the 3-8 MHz range are observed and are suggested to arise from methyl or hydroxyl protons which occur as substituents on the plastosemiquinone cation ring or from amino acid protons hydrogen-bonded to the 1,4-hydroxyl groups. Orientation selection experiments are consistent with these possibilities. D2O/H2O exchange shows that the D(+)/Z(+) site is accessible to solvent. However, the exchange occurs slowly and is not complete even after 72 hours which suggests that the free radicals are functionally isolated from solvent water.