The dark recovery kinetics of the Chl a fluorescence transient (OJIP) after 15 min light adaptation were studied and interpreted with the help of simultaneously measured 820 nm transmission. The kinetics of the changes in the shape of the OJIP transient were related to the kinetics of the qE and qT components of non-photochemical quenching. The dark-relaxation of the qE coincided with a general increase of the fluorescence yield. Light adaptation caused the disappearance of the IP-phase (20-200 ms) of the OJIP-transient. The qT correlated with the recovery of the IP-phase and with a recovery of the re-reduction of P700(+) and oxidized plastocyanin in the 20-200 ms time-range as derived from 820 nm transmission measurements. On the basis of these observations, the qT is interpreted to represent the inactivation kinetics of ferredoxin-NADP(+)-reductase (FNR). The activation state of FNR affects the fluorescence yield via its effect on the electron flow. The qT therefore represents a form of photochemical quenching. Increasing the light intensity of the probe pulse from 1800 to 15000 mumol photons m(-2) s(-1) did not qualitatively change the results. The presented observations imply that in light-adapted leaves, it is not possible to 'close' all reaction centers with a strong light pulse. This supports the hypothesis that in addition to Q(A) a second modulator of the fluorescence yield located on the acceptor side of photosystem II (e.g., the occupancy of the Q(B)-site) is needed to explain these results. Besides, some of our results indicate that in pea leaves state 2 to 1 transitions may contribute to the qI-phase.