Geochemical proxies—such as the isotopic compositions and abundances of redox-sensitive metals, organic biomarkers, and sedimentological indicators—can be used qualitatively and quantitatively to reconstruct marine redox landscapes through Earth history. Isotopic proxies with residence times longer than ocean mixing times, including uranium (238U/235U, commonly denoted as δ238U), are a promising but still developing approach to constrain global changes in redox conditions. Our current understanding of the controls on δ238U variability and associated isotopic fractionations is limited, complicating the interpretation of δ238U records through Earth history. With these gaps in mind, we investigate the major controls on the expression of uranium (U) isotope fractionation within two basins of the Miocene Monterey Formation (ca. 18–6 Ma) that represent contemporaneous deposition under different environmental conditions. Our finding that the isotopic offset in the productive and anoxic Santa Barbara Basin is similar to that in modern euxinic settings suggests that highly productive settings without persistent euxinia can also exert significant leverage on seawater δ238U. Distinct patterns in U concentrations and δ238U between basins demonstrate that local depositional controls can impart a strong influence on the U isotope offset in reducing settings, as has been observed in modern sediment datasets. High productivity and low sedimentation rates are predicted to result in a diagnostic inverse relationship between U enrichment and δ238U, which can partially explain the overall weak negative correlation between [U] and δ238U in the distal Naples Beach section in the Santa Barbara Basin. In a core from a more restricted proximal setting in the San Joaquin Basin, we propose that changes in basin hydrography led to a potential relationship between [U] and δ238U. These interpretations are consistent with the characteristic depositional conditions interpreted independently for each basin (e.g., persistence of anoxia, productivity, sedimentation rates, phosphate content, and basin hydrography) and are supported by modeling results of U enrichment and isotope fractionation under these conditions. In particular, our results suggest that the isotope offset of U in reducing sinks varies and has done so over Earth history and in different paleogeographic locations as a result of local productivity and basin restriction. These variations indicate the need to consider these uncertainties in quantitative models of global redox conditions from δ238U records—and specifically in using δ238U to distinguish between euxinic and oxic redox states without considering other oxygen-depleted environments as part of a continuum of settings that can impact δ238U if spatially widespread. This study also highlights the utility of using different lithologies to evaluate global versus local controls on δ238U records as well as multi-proxy approaches to constrain ancient redox landscapes.