Nitrite reductase, which reduces nitrite to ammonium in a six-electron reaction, was characterized through kinetic analysis of an electron transfer cascade involving photoexcited Photosystem I and ferredoxin. This cascade was studied at physiological pH by flash-absorption spectroscopy. Two different forms of the enzyme were studied: one isolated from spinach leaf and one histidine-tagged recombinant form. When the enzyme is oxidized in the absence of nitrite, single-enzyme reduction leads mostly to siroheme reduction with the leaf enzyme, whereas the siroheme and the [4Fe-4S] cluster are both reduced in equivalent amounts in the recombinant enzyme. When combined with the results of deazaflavin/EDTA photoreduction experiments, these data support a 50 mV negative shift of the siroheme midpoint potential in the recombinant enzyme. Despite this difference, the two forms of the enzyme exhibit similar values for the rate constant of single reduction by reduced ferredoxin (1200 s -1) and for k cat (420-450 electrons per second and per nitrite reductase). When nitrite reductase is initially pre-reduced to the state ferrous siroheme-NO, the fast kinetics of reduction by ferredoxin and the thermodynamics of ferredoxin binding are equivalent to those observed with oxidized nitrite reductase without nitrite. Spectral and kinetic analyses of single reduction of the recombinant enzyme in the ferrous siroheme-NO state by photoreduced ferredoxin reveal that this process leads to reduction of the [4Fe-4S] cluster with little, if any, NO reduction. These data show that the enzyme must wait for the next reduction step before NO undergoes substantial reduction.