Injection of cattle slurry into a grassland soil decreases NH3 volatilisation and increases N utilisation by the sward, but may also increase denitrification losses. Denitrification rates were measured using a soil core incubation technique involving acetylene inhibition, following injection of cattle slurry (67 t ha-1) into a grassland soil. The slurry was injected, either with or without a nitrification inhibitor (DCD), on 8 December 1989. Two-weekly measurements were carried out up to 18 weeks after injection. Compared to the control plot, denitrification rates were significantly higher after slurry injection. Addition of DCD to the slurry almost eliminated this effect. Estimated N-losses during 18 weeks after injection were 0.9 (control), 4.1 (+DCD), and 13.7 (-DCD) kg N ha-1. Denitrification losses were 7% of the injected NH4-N and decreased to 2% of the injected NH4-N when DCD was added. Denitrification could account for about 19% of the difference in apparent recovery of N from slurry injected with and without DCD. The results suggested that considerable amounts of NO3 were lost due to leaching.