A modeling study revealed that the depth of nitric oxide (NO) production in soil is crucial for its flux, while that of nitrous oxide (N2O) is not. To verify this result, laboratory experiments with soil columns classified as Andisol (Hydric Hapludand) were conducted, with changing the depth of urea application, at 0-0.1 or 0.1-0.2 m. All the NO concentration profiles in soil exhibited a sharp peak at each fertilized layer within 5 days of fertilizer application. NO concentration in soil decreased abruptly as the distance from the fertilized layer increased. These findings imply that NO is produced mainly within the fertilized layer, but does not diffuse widely in the soil columns, because of rapid NO uptake within the soil. As a result, the NO flux from soil columns fertilized at 0.1-0.2 m depth over the 48-day study period was reduced to almost the same rate as that of the unfertilized one. The total NO emissions from soil columns unfertilized and fertilized at 0-0.1 and 0.1-0.2 m depth were 0.02, 1.39 (+/- 0.05) and 0.05 (+/- 0.03) kg N ha(-1), respectively, suggesting that NO emission derived from N fertilizer could be reduced to 2% by shifting the depth of fertilizer application by 0.1 m. On the other hand, soil N2O concentration profiles exhibited a gentler peak, because of the lower uptake by soil. N2O fluxes were affected more by the soil conditions, e. g. soil water content, than the distance between fertilized depth and soil surface. The total N2O emissions from soil columns unfertilized and fertilized at 0-0.1 and 0.1-0.2 m were 0.02, 0.16 (+/- 0.03) and 0.25 (+/- 0.04) kg N ha(-1), respectively.