Measurements of some of the main internal N-cycling processes in soil were obtained by labelling the inorganic N pool with the stable isotope of nitrogen (15N). The 15N mean pool dilution technique, combined with other field measurements, enabled gross and net N-mineralization rates to be resolved in grassland soils, which had previously either received fertilizer N (F), or had remained unfertilized (U) for many years. The two soils were subdivided into plots that received N at different time intervals (over 3 weeks), prior to 15N measurements being made. By this novel approach, possible ‘priming’ effects over time were investigated to try to overcome some of the temporal problems of isotopic labelling of soil N (native plus fertilizer) and to identify possible changes in a range of primary N-transformation processes. The results suggested that an overall stimulation of microbially mediated processes occurred with all N treatments, but there were inconsistencies associated with the release of N, both in the timing and the degree to which different processes responded to the application of fertilizer N. The rates of these processes were, however, within the range of previously reported data and the 15N measurements were not adversely affected by the differences in N pools created by the treatments. Thus, the mean pool dilution technique was shown to be applicable to agricultural soils, under conditions relevant to grass swards receiving fertilizer. For example, between the U and F treatments, the size of inorganic N pools increased by five-fold and gross rates of mineralization reached 3.5 and 4.8 µg N g−1 (dry soil) d−1, respectively, but did not vary greatly with the timing of N applications. A correlation (r2 = 0.57) was found between soil respiration (which is relatively simple to measure) and net mineralization (which is more time consuming), suggesting that the former might be used as an indicator of the latter. Although this relationship was stronger in previously unfertilized soils, the similarities found with fertilized soils suggest that this approach could be used to obtain information of wider agronomic value and would, therefore, warrant further work under a range of soil conditions. Copyright © 2002 John Wiley & Sons, Ltd.