AbstractClarifying the fate of different nitrogen (N) species in different pools of terrestrial ecosystems is a prerequisite for a comprehensive understanding of the influence of human activities on the N cycle. Grazing has always been an important way of grassland management for centuries in the temperate grasslands of North China. However, how grazing regimes affect the N fate of urea derived from the livestock in the plant–soil systems of grazed grasslands remains poorly understood. Therefore, an in situ three‐factor (grazing regime, soil depth, and sampling time) 15N‐labeling experiment in a temperate steppe was conducted to answer this question. After 48 days of 15N labeling, the 15N recovered in shoots under no grazing (7.3% ± 1.8%) was approximately 2.3 times of that under rotational (3.2% ± 0.4%) and 2.5 times of that under continuous overgrazing (2.9% ± 0.5%). More 15N was recovered in roots under rotational than continuous overgrazing (19.8% ± 2.4% vs. 10.4% ± 0.5%, respectively), indicating that rotational overgrazing could promote more N retention in the roots. However, the 15N recovered in the soil was lower under continuous (23.7% ± 2.0%) than that of no grazing (42.0% ± 5.6%). Additionally, overgrazing reduced the magnitude of the soil active N pool in microbial biomass N and soluble N relative to no grazing. The grazing regimes would have significantly influenced both the soil and plants. That is to say, grazing regimes have directly impacted plant growth, and subsequently indirectly affected soil properties. Overgrazing often led to excessive vegetation consumption, resulting in decreased soil water content (SWC) and reduced soil organic carbon (SOC), ultimately caused alterations in plant species composition. The retention of 15N within the plant–soil system under continuous overgrazing was notably lower compared to that of no grazing. Continuous overgrazing has led to a shift in the dominant plant species from Leymus chinensis to Stipa grandis, by decreasing the proportion of perennial grasses by 10%, and increasing the annual and biennial plants by 8%. The fate of 15N was also altered in response to the variations in grazing regimes. Consequently, the recovery of 15N within the plant–soil system under continuous overgrazing was significantly lower compared to that of no grazing. In conclusion, overgrazing reduces the recovery of 15N within the plant–soil system in the temperate steppe, and rotational grazing is more preferable over continuous grazing as it could promote higher N retention in grassland ecosystems.