Managed aquifer recharge (MAR) systems can be designed and operated to improve water supply and quality simultaneously by creating favorable conditions for contaminant removal during infiltration through shallow soils. We present results from laboratory flow-through column experiments, using intact soil cores from two MAR sites, elucidating conditions that are favorable to nitrate (NO3) removal via microbial denitrification during infiltration. Experiments focused on quantitative relations between infiltration rate and the presence or absence of a carbon-rich permeable reactive barrier (PRB) on both amounts and rates of nitrate removal during infiltration and associated shifts in microbial ecology. Experiments were conducted using a range of infiltration rates relevant to MAR (0.3–1.4 m/day), with PRBs made of native soil (NS), woodchips (WC) and a 50:50 mixture of woodchips and native soil (MIX). The latter two (carbon-rich) PRB treatments led to statistically significant increases in the amount of nitrate removed by increasing zero-order denitrification rates, both within the PRB materials and in the underlying soil. The highest fraction of nitrate removal occurred at the lowest infiltration rates for all treatments. However, the highest nitrogen mass removal (∆NL) was observed at 0.4–0.7 m/day for both the WC and MIX treatments. In contrast, the maximum ∆NL for the NS treatment was observed at the lowest infiltration rates measured (~0.3 m/day). Further, both carbon-rich PRBs had a substantial impact on the soil microbial ecology in the underlying soil, with lower overall diversity and a greater relative abundance of groups known to degrade carbon and metabolize nitrogen. These results demonstrate that infiltration rates and carbon availability can combine to create favorable conditions for denitrification during infiltration for MAR and show how these factors shape and sustain the microbial community structures responsible for nutrient cycling in associated soils.