One way to increase the amount of carbon sequestered in agricultural land is to convert conventional tillage into no-tillage systems. This greatly affects the location of crop residues in soil. To investigate the impact of the location of residues on soil physical and biological properties and how the interactions between those properties influence the fate of carbon and nitrogen in soil, we did a laboratory experiment with repacked soil in columns. Doubly labelled (CN)-C-13-N-15 oilseed rape residues were incorporated in the 0-10 cm layer or left on the soil surface. The columns were incubated for 9 weeks at 20 degrees C and were submitted to three cycles of drying and wetting, each of them induced by a rain simulator. The location of the residues affected the water dynamics and the distribution of C and N in the soil, which in turn influenced microbial activity and the decomposition rate of the added residues. After 9 weeks of incubation, 18.4 +/- 1.5% of the surface applied residue-C and 54.7 +/- 1.3% of the incorporated residue-C was mineralized. We observed a nitrate accumulation of 10.7 mg N kg(-1) with residues at the soil surface, 3.6 mg N kg(-1) with incorporated residues and 6.3 mg N kg(-1) without addition of fresh organic matter, which entailed net N mineralization in soil under mulch and immobilization of N with residue incorporation compared with the control soil. We concluded that application of oilseed rape residues at the soil surface increased the storage of fresh organic C in soil in the short term, compared with the incorporation treatment, but increased the risk of nitrate leaching.