This study involved the use of soil and suspended growth microcosms to study the variation in groundwater denitrification rates using different substrates. Two gaseous substrates, carbon monoxide and methane, were studied and compared to a common liquid substrate, methanol. Denitrification with carbon monoxide as a substrate was achieved using an acclimated seed of mixed activated sludge and anaerobic digester sludge. Kinetic studies of denitrification using carbon monoxide suggested a strong substrate inhibition effect. The observed maximum denitrification velocity of 0.026 mg N/d-mg VSS occurred at a carbon monoxide partial pressure of 0.10 atmospheres (2.8 mg/ℓ). At higher carbon monoxide partial pressures, denitrification velocities decreased. The denitrification velocities at various carbon monoxide concentrations were described by a modified form of the Haldane substrate inhibition model. The biomass yield using carbon monoxide was 1.1 mg VSS/mg VSS, the maximum specific growth rate was 0.03 mg VSS/d-mg VSS, and the half velocity constant was 26 mg-N/ℓ. Denitrification rates using carbon monoxide as a substrate were much slower than those obtained using methanol, and the cost of carbon monoxide was much higher. Denitrification occurred readily, when methanol was provided as a substrate, in microcosms containing either a clay soil, a sandy soil, or activated sludge. Under the conditions of this study, denitrification was not achieved in clay soil or sandy soil microcosms using methane or carbon monoxide as substrates. Denitrification was not achieved using methane as a substrate with an activated sludge seed. / Ph. D.