Metalloradical EPR signals have been found in intact Photosystem II at cryogenic temperatures. They reflect the light-driven formation of the tyrosine Z radical (Y Z ) in magnetic interaction with the CaMn 4 cluster in a particular S state. These so-called split EPR signals, induced at cryogenic temperatures, provide means to study the otherwise transient Y Z and to probe the S states with EPR spectroscopy. In the S 0 and S 1 states, the respective split signals are induced by illumination of the sample in the visible light range only. In the S 3 state the split EPR signal is induced irrespective of illumination wavelength within the entire 415–900nm range (visible and near-IR region) [Su, J. H., Havelius, K. G. V., Ho, F. M., Han, G., Mamedov, F., and Styring, S. (2007) Biochemistry 46, 10703–10712]. An important question is whether a single mechanism can explain the induction of the Split S 3 signal across the entire wavelength range or whether wavelength-dependent mechanisms are required. In this paper we confirm that the Y Z radical formation in the S 1 state, reflected in the Split S 1 signal, is driven by P680-centered charge separation. The situation in the S 3 state is different. In Photosystem II centers with pre-reduced quinone A (Q A ), where the P680-centered charge separation is blocked, the Split S 3 EPR signal could still be induced in the majority of the Photosystem II centers using both visible and NIR (830nm) light. This shows that P680-centered charge separation is not involved. The amount of oxidized electron donors and reduced electron acceptors (Q A − ) was well correlated after visible light illumination at cryogenic temperatures in the S 1 state. This was not the case in the S 3 state, where the Split S 3 EPR signal was formed in the majority of the centers in a pathway other than P680-centered charge separation. Instead, we propose that one mechanism exists over the entire wavelength interval to drive the formation of the Split S 3 signal. The origin for this, probably involving excitation of one of the Mn ions in the CaMn 4 cluster in Photosystem II, is discussed.