A micrometeorological mass balance technique was used to determine ammonia (NH 3 ) volatilization from pig (Sus scrofa) slurry applied to winter wheat (Triticum aestivum Lam.). The slurry was applied with trail hoses on the soil below the canopy or by a splash plate technique spreading the slurry on both plants and soil. The two application techniques were compared in parallel experiments of 7 d duration on each of nine occasions in the period from April 1993 to June 1995. The loss of NH 3 varied from 4 to 26% of the ammonium (NH 4 + ) in slurry applied with trail hoses and from 11 to 26% when applied with splash plates. Trail hose application reduced NH 3 volatilization by up to 80% compared with the losses from splash plate applied slurry. The greatest reduction was observed when slurry was applied to a tall and dense crop, while the trail hose technique did not reduce losses when slurry was applied to a 10 cm high crop with a leaf area index of 0.3. The decreasing NH 3 volatilization with increasing crop height was due to a reduced wind speed above the slurry surface, promoted slurry infiltration due to increased drying of the top soil and increasing leaf absorption of volatilized NH 3 . Wind speed and air temperature above the canopy and the chemical composition of the slurry had little influence on NH 3 volatilization from trail hose applied slurry. On the contrary, these factors increased NH 3 , volatilization from splash plate applied slurry. The NH 3 volatilized from trail hose applied slurry was absorbed by the wheat plants in rates from 0 to 0.74 g NH 3 -N m -2 leaf surface during a period of 7 d after slurry application. Canopy NH 3 absorption was responsible for up to 25% of the reduction in NH 3 loss when using trail hose application. The wheat plants did not absorb NH 3 during stem elongation in the vegetative growth period.