Combining the remarkable catalytic properties of redox enzymes with highly tunable light absorbing properties of semiconductor nanocrystals enables the light-driven catalysis of complex, multielectron redox reactions. This Article focuses on systems that combine CdS nanorods (NRs) with the MoFe protein of nitrogenase to drive photochemical N2 reduction. We used transient absorption spectroscopy (TAS) to examine the kinetics of electron transfer (ET) from CdS NRs to the MoFe protein. For CdS NRs with dimensions similar to those previously used for photochemical N2 reduction, the rate constant for ET from CdS NRs competes with other electron relaxation processes, such that when a MoFe protein is bound to a NR, about one-half of the photoexcited electrons are delivered to the enzyme. The NR–MoFe protein binding is incomplete with more than one-half of the NRs in solution not having a MoFe protein bound to accept electrons. The quantum efficiency of ET (QEET) in these ensemble samples is similar to previously reported efficiencies of product (NH3 and H2) formation, suggesting that the enzyme utilizes the delivered electrons without major loss pathways. Our analysis suggests that QEET, and therefore the photochemical product formation, is limited at the ensemble level by the NR–MoFe protein binding and at the single-complex level by the competitiveness of ET. We characterized ET kinetics for several CdS NRs samples with varying dimensions and found that for CdS NRs with an average diameter of 4.2 nm the ET efficiency dropped to undetectable levels, defining a maximum NR diameter that should be used to photochemically drive the MoFe protein. The work described here provides insights into the design of systems with increased control of photochemical N2 reduction catalyzed by the MoFe protein of nitrogenase.