Context Organic resources application can increase soil nitrogen (N) sequestion and improve the pigsical structure while preventing soil compaction, yet the impact of vermicompost application on soil macro-and micro-aggregates as related to changes in soil organic N fractions within aggregates has not been studied well, especially in continuous protected vegetable fields with serious deterioration of soil structure. Methods We evaluated the effect of three fertilizer practices on soil dry aggregate size distribution, aggregate stability, particulate organic matter N (PON) and mineral-associated organic matter N (MON) distributions within different particle fractions, as well as their relationships in two soil layers (0-20 cm and 20-40 cm) in a protected tomato (twice a year) continuous cropping system. Treatments included commercial organic fertilizer applied (COF), reduced commercial organic fertilizer combined with vermicompost (RCOF+VC), vermicompost applied only (VC) in a 2-year field trial. Equal amounts of chemical inorganic fertilizers were applied for different treatments. Results Compared with COF, RCOF+VC was not conductive to soil macro-aggregation and aggregate stabilization in 20-40 cm, though larger amounts of organic matter were applied. The disadvantages could be due to reduced PON and MON concentrations within aggregates (being a complete loss in micro-aggregates), which were significantly correlated with soil macro-aggregation and stability characteristics. In comparision, VC had no significant disadvantages on soil macro-aggregation and stability characteristics, while effective in N rehabilitation by preventing degradation of organic N compounds in aggregates, especially in 0-20 cm. For all fertilizer treatments, most PON and MON (average 86.05%) were distributed in the macroaggregate-occluded fractions. The positive regulation of PON and MON from macro-aggregates was larger than that from smaller aggregates, showing the dominant role of organic N fractions within macro-aggregates for the quality of soil agglomeration structure. In all fractioned aggregate sizes, compared with PON, MON completely dominated due to its large mass proportions, regardless of the specific concentrations, demonstrating the MON in macro-aggregates as a primary fate and resevior pool of exogenous organic-derived N. Although with quite different numbers, the PON and MON in macro-aggregates functioned almost equally in promoting soil macro-aggregation. Whereas in the aspect of aggregate stabilization, MON in macro-aggregates played a more pivotal role. A direct regulating role rather than an indirect role through the conversion to MON was found for PON within macro-aggregates. Conclusions We suggest the vermicompost as a economical substitute for commercial organic fertilizer in terms of maintaining soil aggregate structure and the occluded organic N fractions.