Tuning the electron-transfer (ET) steps in redox reactions has proven to be an effective strategy to modulate enzyme catalysis. Here, we use 2-oxoacid:ferredoxin oxidoreductase (OFOR) as a model system to study the two-electron reduction of CO 2 and to evaluate the effect of intermolecular ET on catalysis by systematically engineering the reduction potential (E m ) of the electron donor, ferredoxin (Fd), aiming to maximize catalytic CO 2 reduction. Based on the newly determined crystal structure for Fd1 from Hydrogenobacter thermophilus (HtFd1), we expanded the E m of HtFd1 (−485 mV, vs the standard hydrogen electrode) to a range from −440 to −555 mV through changes in the hydrogen bonding network and solvent exposure of the [4Fe− 4S] cluster, with −555 mV being the lowest potential observed for a monocluster [4Fe−4S] Fd. Specifically, replacement of a polar residue (serine) with a non-polar residue (alanine or valine) at the +2 position with respect to the C4 of the C1XXC2XXC3X n C4 [4Fe−4S] cluster-ligating motif abolished an extensive hydrogen bonding network and decreased E m by 40 mV; the introduction of a non-ligating "fifth cysteine" at the +4 position of C4 decreased E m by 60 mV; and increasing the solvent exposure of the [4Fe−4S] cluster by replacement of a bulky residue (leucine) with glycine at the −1 position of C3 increased E m by 45 mV. We constructed an Fd library of 24 Fds with a "potential gradient", ranging from −378 to −555 mV, and evaluated the impact of the Fd potential on OFOR activity. Here, we found that decreasing the Fd potential and therefore providing a larger driving force are beneficial for CO 2 reduction, but only to a point. With the HtFd1_T13I_S64A variant (−515 mV), we were able to achieve a maximum turnover frequency of 81.3 min −1 (or a specific activity of 625 nmol min −1 mg −1 ) in CO 2 reduction.