Land application of liquid manures is a major source of atmospheric ammonia. The presence of crop residues on the soil surface usually increases emissions by retarding slurry infiltration, whereas incorporation of slurry into soil reduces emissions. Our objective was to quantify the relative reduction in NH3 volatilization resulting from the soil incorporation of pig slurry (PS) applied on rape (Brassica napus) residues under autumn conditions in Quebec, Canada. Pig slurry was applied at 7.4 litre m-2 on six plots covered by rape crop residues. Slurry and residues were incorporated in the top 5 cm of soil (INCORP) in half of the plots, while the other half were left untouched (SURF). Ammonia volatilization was measured following application for 10 days using wind tunnels. Soil NH4+ and NO3- contents, pH, moisture and temperature were also monitored to explain variations in NH3 fluxes. Soil NH4+-N in the surface soil was lower than expected shortly after slurry application, maybe as a result of fixation by clays or interception by crop residues. The volatilization of NH3 was higher (P<0.05) on SURF plots than on INCORP plots in 20 of the 26 measuring periods, with total NH3 losses being five times greater in the former. Cumulated emissions during the first 11 h accounted for the 60 and 53% of total NH3 emissions for the SURF and INCORP plots, respectively. Results confirm that a large fraction of the NH3 volatilization from slurry application on rape residues can be greatly reduced if the slurry and crop residues are incorporated into the soil immediately after slurry application. Despite significant reduction (80%) of NH3 volatilization in INCORP compared with SURF plots, no difference was found in soil mineral N between treatments, suggesting that other processes such as N mineralization or denitrification were more active in INCORP plots.