Abstract-The delayed light emission decay rate (up to 120~s) and the rise in chlorophyll a fluorescence yield (from 3 to 35 ys) in isolated chloroplasts from several species, following a saturating 10 ns flash, are temperature independent in the CL35"C range. However, delayed light in the 120-340ps range is temperature dependent. Arrhenius plots of the exponential decay constants are: (a) linear for lettuce and pea chloroplasts but discontinuous for bush bean (12-17°C) and spinach (12-20°C) chloroplasts; (b) unaffected by 3-(3,4 dich1orophenyl)l ,1-dimethylurea (inhibitor of electron flow), gramicidin D (which eliminates light-induced membrane potential) and glutaraldehyde fixation (which stops gross structural changes).The discontinuities, noted above for bush bean and spinach chloroplasts, are correlated with abrupt changes in (a) the thylakoid membrane lipid fluidity (monitored by EPR spectra of 12 nixtroxide stearate, 12 NS) and (b) the fluidity of extracted lipids (monitored by differential calorimetry and EPR spectra of 12 NS). However, no such discontinuity was observed in (a) chlorophyll a fluorescence intensity of thylakoids and (b) fluorescence of tryptophan residues of delipidated chloroplasts.Microsecond delayed light is linearly dependent on light intensity at flash intensities as low as one quantum per 2 x lo4 chlorophyll molecules. We suggest that this delayed light could originate from a one quantum process in agreement with the hypothesis that recombination of primary charges leads to this light emission. A working hypothesis for the energy levels of Photosystem I1 componcnts is proposed involving a charge stabilization step on the primary acceptor side, which is in a lipid environment.Finally, the redox potential of P680 (the reaction center for chlorophyll of system 11) is calculated to he close to 1.&1.3V.