Parts of the South Platte River alluvial aquifer in northeastern Colorado are underlain by the Pierre Shale, a marine deposit of Late Cretaceous age that is <1000 m thick. Ground water in the aquifer is contaminated with NO₃^‐, and the shale contains abundant potential electron donors for denitrification in the forms of organic carbon and sulfide minerals. Nested piezometers were sampled, pore water was squeezed from cores of shale, and an injection test was conducted to determine if denitrification in the shale was a sink for alluvial NO₃⁻ and to measure denitrification rates in the shale. Measured values of NO₃⁻, N₂, NH₄⁺, δ¹⁵N[NO₃⁻], δ¹⁵N[N₂], and δ¹⁵N[NH₄⁺] in the alluvial and shale pore water indicated that denitrification in the shale was a sink for alluvial NO₃⁻. Chemical gradients, reaction rate constants, and hydraulic head data indicated that denitrification in the shale was limited by the slow rate of NO₃⁻ transport (possibly by diffusion) into the shale. The apparent in situ first‐order rate constant for denitrification in the shale based on diffusion calculations was of the order of 0.04–0.4 yr⁻¹, whereas the potential rate constant in the shale based on injection tests was of the order of 60 yr⁻¹. Chemical data and mass balance calculations indicate that organic carbon was the primary electron donor for denitrification in the shale during the injection test, and ferrous iron was a minor electron donor in the process. Flux calculations for the conditions encountered at the site indicate that denitrification in the shale could remove only a small fraction of the annual agricultural NO₃^‐input to the alluvial aquifer. However, the relatively large potential first‐order rate constant for denitrification in the shale indicated that the percentage of NO₃⁻ uptake by the shale could be considerably larger in areas where NO₃⁻ advection.