The way misses happen in oxygen evolution is subject to debate (Govindjee et al. 1985). We recently observed a linear lowering of the miss probability with the flash number (Meunier and Popovic 1989). Therefore, we investigated in Dunaliella tertiolecta the link between the average miss probability and the redox state of plastoquinone after n flashes. The effect of flashes was to oxidize the plastoquinone pool; we found that the oxidation of plastoquinone highly correlated (linear regression: R (2)=0.996) with the lowering of the miss probability. The flash frequency was found to affect both the miss probability and the redox state of plastoquinone. When pre-flashes were given using a high flash frequency (10 Hz), the plastoquinone pool was oxidized and misses were low; however, if long dark intervals between flashes were used, the oxidizing effect of flashes was lost and the misses were high. We could not explain our results by assuming equal misses over all S-states; but unequal misses, caused by deactivations, were coherent with our results. We deduced that chlororespiration was responsible for the reduction of plastoquinone in the dark interval between flashes. We compared oxygen evolution with and without benzoquinone, using a low flash frequency (0.5 Hz) for maximum misses. Benzoquinone lowered the misses from 34% to 3%, and raised the amplitude of oxygen evolution by more than a factor of two (2). From this we deduced that the charge carrier "C" postulated to explain misses (Lavorel and Maison-Peteri 1983) did not account for more than 3% of miss probability in Dunaliella tertiolecta. These results indicate that the misses in oxygen evolution are controlled by the redox state of plastoquinone, through deactivations.