The microplot method was used to compare soils under different long-term (19 years) fertilization treatments i.e. NF (no fertilizer); NPK (inorganic NPK fertilizer) and MNPK (manure plus inorganic NPK fertilizer) in retention, crop uptake, and leaching of applied N. Results show that application of N significantly increased mineral N in the soil of Treatment NF, but had no such effect in the soils of Treatments NPK and MNPK. However, it increased soil microbial biomass N (SMBN) by 4.3 and 0.8 times in the soils of Treatments NPK and MNPK at the elongating stage of wheat, but did not in the soil of Treatment NF. From the elongating stage to the flowering stage, soil microbial biomass N in the soils of Treatments NPK and MNPK dropped by 51% and 56%. In Treatments NPK and MNPK, the crop of wheat used up 36% and 45%, respectively, of applied N, while in Treatment NF, wheat rarely absorbed any of the applied N, of which 34% was used up by the subsequent maize. By the time when wheat was harvested more than 50% of the applied N had been leached into the soil layers below 30 cm in Treatment NF, and a significant portion into the 30～50 cm soil layer in Treatment NPK, but little down the soil profile in Treatment MNPK. The findings demonstrate that long-term combined application of manure and NPK is an effective way of fertilization management that could strengthen N buffer capacity of the soil, harmonize the relationship between soil N retention and crop N uptake demand, improve N use efficiency and minimize the risk of environment pollution.