Dedicated biomass crops are required for future bioenergy production. However, the effects of large-scale land use change (LUC) from traditional annual crops, such as corn–soybean rotations to the perennial grasses (PGs) switchgrass and miscanthus, on soil microbial community functioning is largely unknown. Specifically, ecologically significant denitrifying communities, which regulate N 2 O production and consumption in soils, may respond differently to LUC due to differences in carbon (C) and nitrogen (N) inputs between crop types and management systems. Our objective was to quantify bacterial denitrifying gene abundances as influenced by corn–soybean crop production compared to PG biomass production. A field trial was established in 2008 at the Elora Research Station in Ontario, Canada ( n = 30), with miscanthus and switchgrass grown alongside corn–soybean rotations at different N rates (0 and 160 kg N ha −1 ) and biomass harvest dates within PG plots. Soil was collected on four dates from 2011 to 2012 and quantitative PCR was used to enumerate the total bacterial community (16S rRNA) and communities of bacterial denitrifiers by targeting nitrite reductase ( nirS ) and N 2 O reductase ( nosZ ) genes. Miscanthus produced significantly larger yields and supported larger nosZ denitrifying communities than corn–soybean rotations regardless of management, indicating large-scale LUC from corn–soybean to miscanthus may be suitable in variable Ontario climatic conditions and under varied management, while potentially mitigating soil N 2 O emissions. Harvesting switchgrass in the spring decreased yields in N-fertilized plots, but did not affect gene abundances. Standing miscanthus overwinter resulted in higher 16S rRNA and nirS gene copies than in fall-harvested crops. However, the size of the total (16S rRNA) and denitrifying bacterial communities changed differently over time and in response to LUC, indicating varying controls on these communities.