Spatial and temporal variation in N‐gas loss (N2O evolution with C2H2), denitrifying enzyme activity (DEA), and denitrifier biomass (as determined by most‐probable‐number [MPN] counts) was measured in two agricultural soils in central Kentucky in order to characterize the extent to which denitrifying enzymes are synthesized and expressed, and how environmental conditions influence the in situ growth and activity of denitrifying bacteria. Rates of N‐gas loss from intact cores ranged from 0 to 300 µg N m−2 h−1. Nitrogen gas loss was strongly seasonal, with both soils having the most frequent activity in the spring. Nitrogen gas loss was highly variable on a spatial scale with replicate measurements exhibiting coefficients of variation between 74 and 268%. Denitrifying enzyme activity in both soils ranged from 0.1 to 119 ng N2O‐N g−1 soil min−1 and varied spatially and temporally, but did not exhibit the same seasonal pattern as N‐gas loss. Increases in DEA often followed increases in soil moisture and were commonly associated with increases in soil respiration. Denitrifier populations ranged from 3 × 103 to 1 × 106 cells g−1 dry soil but did not show the same temporal pattern as N‐gas evolution.When the entire data set was considered, none of the parameters measured, alone or in combination, adequately accounted for variation in N‐gas loss from either soil. When mean values of the parameters measured at each sampling date were considered, however, multiple‐regression models accounted for 74 to 91% of the variation in mean denitrification rate. Denitrifying enzyme activity and MPN population measurements were not good predictors of denitrification rate because N‐gas loss was not always tightly coupled to the synthesis of denitrifying enzymes or population growth.