An alternative approach to quantification of the contribution of non-QB-reducing centers to Chl a fluorescence induction curve is proposed. The experimental protocol consists of a far-red pre-illumination followed by a strong red pulse to determine the fluorescence rise kinetics. The far-red pre-illumination induces an increase in the initial fluorescence level (F(25 micros)) that saturates at low light intensities indicating that no light intensity-dependent accumulation of QA - occurs. Far-red light-dose response curves for the F(25 micros)-increase give no indication of superimposed period-4 oscillations. F(25 micros)-dark-adaptation kinetics following a far-red pre-pulse, reveal two components: a faster one with a half-time of a few seconds and a slower component with a half-time of around 100 s. The faster phase is due to the non-QB-reducing centers that re-open by recombination between QA - and the S-states on the donor side. The slower phase is due to the recombination between QB - and the donor side in active PS II reaction centers. The pre-illumination-induced increase of the F(25 micros)-level represents about 4-5% of the variable fluorescence for pea leaves ( approximately 2.5% equilibrium effect and 1.8-3.0% non-QB-reducing centers). For the other plant species tested these values were very similar. The implications of these values will be discussed.