Monovalent anions, such as formate and particularly bicarbonate, can control the rate of electron flow between Photosystem I and II. There are two explanations for this effect that differ in the residency state of a high-affinity binding site for bicarbonate when normal (rapid) electron flow occurs. One explanation, the 'bound-bicarbonate' hypothesis, requires that the site be filled with bicarbonate for high rates of electron flow. A second explanation, the ‘inhibitory-anion’ hypothesis, requires that the site either be filled with bicarbonate or be entirely free of any inhibitory monovalent anion. Electron flow in maize thylakoids was monitored by rates of oxygen evolution with dichlorobenzoquinone as an electron acceptor. The amount of bicarbonate bound to the high-affinity site in maize thylakoids was determined by measuring, the amount of bicarbonate displaced by injection of formate. The following results were obtained. Thylakoids with high-affinity binding sites devoid of bicarbonate and other monovalent anions have high rates of electron flow. The dissociation constant for bicarbonate at the high-affinity site is 40 ± 11 μM; there are 2.4 ± 0.6 bicarbonate bound and 1.8 ± 0.3 atrazine bound per Photosystem II reaction center active in charge separation; in maize thylakoids 94% of separated charge in Photosystem II is used in oxygen evolution. Uncharacterized respiratory processes occur in spinach and pea thylakoid preparations. It is concluded that the inhibitory-anion hypothesis is correct. Also, bicarbonate might be present for high rates of electron flow in vivo, in which case it binds to its high-affinity site and excludes other monovalent anions that are inhibitory.