An extensive picosecond fluorescence study has been performed on isolated D1-D2 reaction centers (RCs) from photosystem (PS) II at room temperature as a function of excitation and emission wavelength with a time resolution of about 2 ps. The study comprizes more than 50 excitation/emission wavelength pairs. Global analysis for individual excitation wavelengths and, alternatively, for the entire excitation/emission wavelength matrix has been performed. These analyses show that over a time range of 2 ns five exponentials are required to adequately describe the data. The shortest lifetime component, whose amplitude dominates the excited state decay for excitation in the red part of the absorption spectrum, is approximate to 3 ps. Further lifetime components fall into the 6-12, 20-30, and 60-130 ps and the several nanosecond ranges. In the combined global analysis at least five lifetime components (400 ps fit window), again with approximate to 3 ps as the shortest one (again dominant in amplitude for long excitation wavelengths), and most likely even six components are required to describe the data. The dominance and necessity of the approximate to 3 ps component is demonstrated also by exhaustive search error analysis on the data in the global mode. Kinetic target analysis on models of various complexity has been performed, again on data sets from individual excitation wavelengths and on the combined data set. Such analyses have been performed for the first time for the D1-D2 kinetics, and they allow an assignment of the origin of the various lifetime components. The target analyses yielded realistic species associated spectra (SAS), as expected for Chi excited states and also the rate constants for the energy transfer and charge separation processes. The results show that (i) the majority of the effective primary charge separation is associated with the approximate to 3 ps lifetime, (ii) three pools of external Chls, transferring their energy relatively slowly to the RC core, are attached to the D1-D2 RC, (iii) the three Chls have energy transfer times in the range of 6-30 ps, and (iv) a approximate to 50 ps component is associated with a transition between different radical pairs. We exclude on the basis of these data that the primary charge separation could occur primarily in 21 ps, as proposed by other authors (Durrant; et al. Biochemistry 1993, 32, 8259-8267). In all models, quite independent of their complexity, the effective primary charge separation rate (from the equilibrated RC core) is in the range of 100-150 ns(-1). If one accounts for the distribution of the excited state in the RC core over various chromophores an intrinsic charge separation rate constant from P680(double dagger) to the primary radical pair can be estimated to be approximate to 360 ns(-1) corresponding to an intrinsic charge separation lifetime of approximate to 2.7 ps, in very good agreement with our previous extrapolation based on the kinetics of intact PS II core particles (Schatz; et al. Biophys. J. 1988, 54, 397-405). It is concluded that the intrinsic primary charge separation in isolated D1-D2 reaction centers has about the same rate constant as in intact photosystem II.