The pheophytin a Q(x) spectral region of the isolated photosystem II reaction center was investigated at 7 K using femtosecond transient absorption spectroscopy. At this temperature, uphill energy transfer, which greatly complicates the interpretation of the kinetics at or near room temperature, should be essentially shut off. Low-energy (similar to 100 nJ) pulses at 661 and 683 nm were used to excite the short-wavelength and long-wavelength sides of the composite Q(y), band, providing preferential excitation of the accessory pigment pool and P680, respectively. The data analysis uses a background subtraction technique developed earlier (Greenfield et al. J. Phys. Chem. B 1997, 101, 2251-2255) to remove the kinetic components of the data that are due to the large time-dependent changes in the background that are present in this spectral legion. The instantaneous amplitude of the bleach of the pheophytin a Q(x), band with 683 nm excitation is roughly two-thirds of its final amplitude, providing strong evidence of a multimer description of the reaction center core. The subsequent growth of the bleach shows biphasic kinetics, similar to our earlier results at 278 K. The rate constant of the faster component is (5 ps)(-1) for 683 nm excitation (a factor of almost two faster than at 278 K), and represents the intrinsic rate constant for charge separation. The bleach growth with 661 nm excitation is also biphasic; however, the faster component appears to be a composite of a (5 ps)(-1) component corresponding to charge separation following subpicosecond energy transfer to the long-wavelength pigments and a roughly (22 ps)(-1) component corresponding to charge separation limited by slow energy transfer, The combined quantum yield for these two energy transfer processes is near unity. For both excitation wavelengths, there is also a roughly ( 100 ps)(-1) component to the bleach growth. Exposure to high excitation energies (greater than or equal to 1 mu J) at 683 nm results in a substantial permanent loss of ground-state absorption at 680 nm. The transient behavior of these degraded samples is also examined and is consistent with the (5 ps)(-1) rate constant for charge separation. Our results are compared to other low-temperature transient absorption and hole burning studies, as well as to our 278 K results.